ELUCIDATION	
  OF	
  ILBP	
  FAMILY	
  FOLDING	
  PATHWAY	
  AND	
  STUDY	
  OF	
  REENGINEERING	
  THEM	
  AS	
  
FLUORESCENT	
  PROTEIN	
  TAGS	
  VIA	
  STRUCTURAL	
  ANALYSIS	
  
	
  
By	
  
	
  Zahra	
  Assar	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

	
  
A	
  DISSERTATION	
  	
  
Submitted	
  to	
  
	
  Michigan	
  State	
  University	
  	
  
in	
  partial	
  fulfillment	
  of	
  the	
  requirements	
  	
  
for	
  the	
  degree	
  of	
  	
  
	
  
Chemistry	
  -­‐	
  Doctor	
  of	
  Philosophy	
  	
  
	
  
2018	
  

	
  

	
  

ABSTRACT	
  
ELUCIDATION	
  OF	
  ILBP	
  FAMILY	
  FOLDING	
  PATHWAY	
  AND	
  STUDY	
  OF	
  REENGINEERING	
  THEM	
  AS	
  
FLUORESCENT	
  PROTEIN	
  TAGS	
  VIA	
  STRUCTURAL	
  ANALYSIS	
  
	
  
By	
  
	
  Zahra	
  Assar	
  
The	
  intracellular	
  lipid	
  binding	
  proteins	
  (iLBP)	
  family	
  are	
  found	
  in	
  the	
  cells	
  of	
  mammals,	
  
birds,	
   fish,	
   amphibians	
   and	
   reptiles.	
   They	
   function	
   to	
   shuttle	
   large	
   insoluble	
   hydrophobic	
  
molecules,	
  including	
  retinal,	
  various	
  long	
  chain	
  fatty	
  acids	
  and	
  etc.	
  throughout	
  the	
  cell	
  around	
  
the	
   cytosol	
   and	
   nucleus.	
   The	
   combination	
   of	
   their	
   small	
   size	
   and	
   relatively	
   large	
   binding	
   pocket	
  
make	
  them	
  suitable	
  templates	
  in	
  a	
  variety	
  of	
  protein	
  design	
  applications,	
  including	
  the	
  study	
  of	
  
an	
  innovative	
  class	
  of	
  fluorescent	
  proteins.	
  To	
  pursue	
  our	
  goals,	
  we	
  used	
  human	
  Cellular	
  Retinol	
  
Binding	
   Protein	
   II	
   (hCRBPII).	
   We	
   were	
   the	
   first	
   group	
   to	
   achieve	
   the	
   structure	
   of	
   an	
   all-­‐trans-­‐
retinal	
  and	
  the	
  first	
  bonafide	
  structure	
  of	
  retinol-­‐bound	
  hCRBPII.	
  	
  
In	
   the	
   course	
   of	
   these	
   studies,	
   we	
   have	
   discovered	
   hCRBPII	
   is	
   surprisingly	
   capable	
   of	
  
folding	
  as	
  domain	
  swapped	
  dimer	
  (DS),	
  with	
  single	
  mutations	
  able	
  to	
  shift	
  the	
  folding	
  product	
  
from	
  monomer	
  to	
  dimer.	
  Structural	
  analysis	
  of	
  both	
  wild	
  type	
  and	
  	
  multiple	
  mutant	
  DS	
  dimers	
  
led	
  us	
  to	
  remarkable	
  hypotheses	
  regarding	
  mechanism	
  of	
  this	
  phenomenon,	
  which	
  is	
  different	
  
from	
   the	
   previous	
   studies	
   on	
   this	
   family.	
   We	
   proposed	
   that	
   the	
   N-­‐terminal	
   and	
   C-­‐terminal	
  
halves	
   of	
   hCRBPII	
   are	
   capable	
   of	
   at	
   least	
   partially	
   folding	
   independently,	
   to	
   form	
   “open	
  
monomer.”	
   	
   The	
   dimer/monomer	
   ratio	
   depends	
   on	
   the	
   relative	
   rates	
   of	
   dimerization	
   of	
   the	
  
open	
  monomers,	
  versus	
  closing	
  of	
  the	
  two	
  halves	
  together	
  to	
  form	
  the	
  “closed	
  monomer”.	
   In	
  
addition,	
  by	
  comparing	
  structures	
  of	
  holo	
  hCRBPII	
  DS	
  dimer	
  variants,	
  we	
  identified	
  an	
  extremely	
  
large	
  change	
  in	
  the	
  relative	
   orientation	
  of	
   the	
  two	
  domains	
   upon	
   ligand	
   binding	
  in	
   dimers.	
  This	
  

suggests	
   the	
   possibility	
   that	
   iLBP	
   domain	
   swapped	
   dimers	
   could	
   be	
   allosterically	
   regulated	
  
forms	
  of	
  these	
  proteins,	
  at	
  least	
  in	
  some	
  cases.	
  Fatty	
  acid	
  binding	
  protein	
  5	
  (FABP5),	
  another	
  
member	
  of	
  iLBPs,	
  has	
  also	
  been	
  reported	
  to	
  forms	
  a	
  very	
  similar	
  DS	
  dimer,	
  which	
  makes	
  it	
  likely	
  
that	
  other	
  family	
  members	
  could	
  also	
  form	
  DS	
  dimers,	
  and	
  have	
  physiological	
  importance	
  for	
  
some	
  members	
  of	
  the	
  family.	
  	
  
As	
  mentioned,	
  a	
  new	
  class	
  of	
  fluorogenic	
  proteins	
  was	
  created	
  by	
  binding	
  fluorophore	
  
aldehydes	
  in	
  the	
  binding	
  pocket	
  of	
  hCRBPII	
  via	
  protonated	
  Schiff	
  base	
  (PSB)	
  formation.	
  In	
  this	
  
new	
   system,	
   emission	
   of	
   the	
   designed	
   solvatochromic	
   fluorophore	
   is	
   flexible	
   based	
   on	
   the	
  
polarity	
  of	
  the	
  environment;	
  therefore	
  multicolor	
  probes	
  can	
  be	
  developed.	
  More	
  importantly,	
  
absorption/emission	
  wavelengths	
  can	
  be	
  tuned	
  therefor;	
  nonspecific	
  labeling	
  and	
  background	
  
fluorescence	
  can	
  be	
  reduced.	
  By	
  now,	
  the	
  absorption	
  maxima	
  are	
  tuned	
  from	
  501nm	
  to	
  705nm	
  
and	
  emission	
  maxima	
  from	
  613	
  nm	
  to	
  744	
  nm.	
  Covering	
  both	
  the	
  red	
  and	
  far-­‐red	
  fluorescence	
  
wavelength	
  regimes.	
  	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
Copyright	
  by	
  	
  
ZAHRA	
  ASSAR	
  
2018	
  

	
  
	
  
	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
This	
  Dissertation	
  is	
  lovingly	
  dedicated	
  to	
  my	
  grandmother,	
  Maman	
  joon.	
  

	
  

v	
  
	
  

ACKNOWLEDGMENTS	
  
	
  
I	
  am	
  very	
  much	
  grateful	
  to	
  my	
  PhD	
  advisor,	
  Professor	
  James	
  H.	
  Geiger,	
  for	
  his	
  support	
  
and	
  guidance	
  throughout	
  my	
  graduate	
  work.	
  He	
  taught	
  me	
  how	
  to	
  think	
  and	
  design	
  strategies	
  
as	
  a	
  scientist.	
  	
  He	
  also	
  trained	
  me	
  to	
  think	
  independently	
  and	
  to	
  be	
  more	
  patient	
  in	
  science	
  
because	
  “Good	
  things	
  do	
  not	
  come	
  easy.	
  The	
  roak	
  is	
  lined	
  with	
  pitfalls”.	
  Desi	
  Arnaz.	
  
	
  Jim	
  always	
  supported	
  me	
  and	
  my	
  thoughts	
  about	
  my	
  projects.	
  Not	
  only	
  he	
  is	
  a	
  great	
  advisor,	
  
also	
   he	
   is	
   very	
   friendly	
   approachable	
   person.	
   I	
   am	
   so	
   glad	
   and	
   honored	
   to	
   have	
   him	
   as	
   my	
  
advisor.	
   Thank	
   you	
   Jim,	
   for	
   letting	
   me	
   to	
   grow,	
   even	
   start	
   my	
   professional	
   career	
   before	
   I	
  
defend	
  and	
  thank	
  you	
  for	
  everything.	
  
	
  

I	
   would	
   like	
   to	
   thank	
   my	
   second	
   reader	
   and	
   committee	
   member,	
   Professor	
   Babak	
  

Borhan.	
  I	
  still	
  remember	
  August	
  2012	
  when	
  I	
  first	
  came	
  to	
  US,	
  I	
  was	
  not	
  sure	
  about	
  my	
  graduate	
  
research	
   field	
   but	
   Babak	
   guided	
   me	
   very	
   well.	
   	
   Also	
   during	
   biomeetings,	
   He	
   always	
   pointed	
  
important	
   discussions	
   and	
   suggested	
   critical	
   experiments	
   which	
   they	
   all	
   helped	
   me	
   to	
  
understand	
  my	
  research	
  better.	
  Whenever	
  I	
  needed	
  help	
  he	
  was	
  always	
  available	
  for	
  me.	
  Thank	
  
you	
  Babak	
  for	
  everything.	
  
I	
   would	
   like	
   to	
   appreciate	
   Dr.	
   William	
   Henry	
   for	
   being	
   on	
   my	
   committee	
   and	
   reading	
   my	
  
dissertation.	
  I	
  received	
  some	
  great	
  advice	
  from	
  him	
  regarding	
  my	
  research.	
  In	
  addition,	
  thank	
  
you	
  to	
  Professor	
  Gary	
  Blanchard	
  for	
  also	
  reading	
  my	
  dissertation.	
  
I	
  would	
  like	
  to	
  thank	
  Dr.	
  Zahra	
  (Rafida)	
  Nossoni	
  and	
  Dr.	
  Wenjing	
  Wang	
  for	
  opening	
  a	
  path	
  for	
  
me	
   to	
   pursue	
   my	
   PhD.	
   I	
   am	
   happy	
   that	
   I	
   had	
   Rafida	
   right	
   when	
   I	
   joined	
   the	
   Geiger	
   lab.	
   She	
  

vi	
  
	
  

helped	
  me	
  in	
  my	
  first	
  year	
  a	
  lot	
  and	
  she	
  was	
  always	
  available	
  especially	
  when	
  I	
  needed	
  help	
  in	
  
data	
  collection.	
  She	
  is	
  a	
  great	
  sister-­‐in-­‐law	
  and	
  like	
  a	
  sister	
  that	
  I	
  have	
  never	
  had.	
  
I	
   also	
   would	
   like	
   to	
   thank	
   Dr.	
   Chrysoula	
   Vasileiou	
   for	
   her	
   help.	
   She	
   always	
   was	
   available	
   and	
  
helpful	
  to	
  me	
  with	
  her	
  ideas	
  and	
  great	
  discussions	
  in	
  our	
  biomeetings.	
  	
  Thank	
  You	
  Chrysoula.	
  
I	
   am	
   heartily	
   thankful	
   to	
   our	
   awesome	
   group	
   members,	
   all	
   current	
   and	
   former	
   members	
   in	
  
Geiger’s	
  group	
  for	
  being	
  friendly	
  and	
  helpful	
  to	
  me.	
  Specifically,	
  thank	
  you	
  Dr.	
  Remie	
  Fawaz,	
  Dr.	
  
Camille	
  Watson	
  and	
  Dr.	
  Meisam	
  Nosrati,	
  I	
  still	
  remember	
  my	
  first	
  Argonne	
  trip	
  and	
  how	
  much	
  
Meisam	
   helped	
   me	
   during	
   my	
   first	
   time	
   data	
   collection.	
   	
   Thank	
   you	
   Hadi	
   Nayebi	
   and	
   Alireza	
  
Ghanbarpour	
   for	
   being	
   great	
   lab	
   members,	
   I	
   always	
   enjoyed	
   our	
   scientific	
   discussions	
   and	
  
chats.	
   Also	
   thank	
   you	
   Nona	
   Ehyaei	
   for	
   always	
   cheering	
   me	
   up	
   with	
   your	
   energy	
   and	
   helping	
   me	
  
whenever	
  I	
  could	
  not	
  drive	
  to	
  East	
  Lansing.	
  Thank	
  you	
  Lindsey	
  Gilbert	
  and	
  Janice	
  Chiou,	
  Kevin	
  
Kramer	
   and	
   Colin	
   McCornack.	
   	
   I	
   would	
   like	
   to	
   thank	
   the	
   bio	
   people	
   in	
   Professor	
   Borhan’s	
   group	
  
Dr.	
  Tetyana	
  Berbasova,	
  Dr.	
  Ipek	
  Yapici,	
  Wei	
  Sheng	
  and	
  especially	
  Dr.	
  Elizabeth	
  Santos.	
  Liz	
  was	
  
the	
  best	
  roommate	
  to	
  live	
  and	
  great	
  scientist	
  to	
  work	
  with.	
  	
  
I	
  would	
  like	
  to	
  thank	
  LS-­‐CAT	
  staff	
  at	
  the	
  Argonne	
  National	
  Laboratory	
  for	
  all	
  of	
  their	
  help	
  during	
  
data	
   collection.	
   In	
   addition,	
   I	
   am	
   thankful	
   to	
   my	
   new	
   scientific	
   family	
   of	
   Cayman	
   Chemical	
  
Company,	
  Dr.	
  Adam	
  Stein	
  and	
  Melissa	
  Holt.	
  I	
  am	
  very	
  happy	
  to	
  work	
  with	
  them	
  as	
  a	
  member	
  of	
  
structural	
  biology	
  team.	
  	
  
I	
  also	
  appreciate	
  my	
  dear	
  friends	
  who	
  supported	
  me	
  on	
  this	
  path.	
  	
  
Finally,	
   I	
   am	
   grateful	
   to	
   my	
   beloved	
   family.	
   My	
   parents,	
   brother	
   and	
   my	
   husband	
   Dr.	
   Farid	
  
Nossoni,	
   without	
   them,	
   their	
   love	
   and	
   support,	
   it	
   would	
   not	
   be	
   possible	
   for	
   me	
   to	
   finish	
   this	
  
stage.	
  

vii	
  

Farid	
  always	
  taught	
  me	
  to	
  be	
  positive.	
  I	
  am	
  thankful	
  to	
  have	
  him	
  in	
  my	
  life,	
  for	
  inspiring	
  me	
  and	
  
supporting	
  me	
  to	
  pursue	
  my	
  path.	
  
And	
  again	
  I	
  would	
  like	
  to	
  say	
  thank	
  you	
  to	
  all	
  of	
  the	
  amazing	
  people	
  who	
  supported	
  and	
  helped	
  
me.	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
‘	
  
	
  
	
  
	
  
	
  
viii	
  

TABLE	
  OF	
  CONTENTS	
  
	
  
LIST	
  OF	
  TABLES………………………………………………………………………………………………………………………...xii	
  
LIST	
  OF	
  FIGURES……………………………………………………………………………………………………………………...xiii	
  
KEY	
  TO	
  SYMBOLS	
  AND	
  ABBREVIATIONS………………………………………………………………………………....xxii	
  
Chapter	
  I:	
  Structural	
  Studies	
  of	
  Human	
  Cellular	
  Retinol	
  Binding	
  Protein	
  II	
  (hCRBPII)	
  bound	
  to	
  
Retinol	
  and	
  Retinal.	
  ................................................................................................................	
  1	
  
I-­‐1	
  Introduction	
  ..........................................................................................................................	
  1	
  
I-­‐2	
  The	
  affinity	
  of	
  hCRBPII	
  for	
  retinol	
  and	
  retinal	
  .......................................................................	
  4	
  
I-­‐3	
  Structure	
  of	
  the	
  retinal-­‐bound	
  hCRBPII	
  complex	
  ..................................................................	
  5	
  
I-­‐4	
  Wavelength-­‐dependent	
  damage	
  of	
  the	
  retinol	
  in	
  hCRBPII	
  ...................................................	
  8	
  
I-­‐5	
  The	
  structural	
  difference	
  between	
  retinol-­‐bound	
  hCRBPII	
  and	
  retinal-­‐bound.	
  ..................	
  13	
  
I-­‐6	
  Ligand	
  Binding	
  in	
  CRBPI	
  versus	
  CRBPII	
  ................................................................................	
  15	
  
I-­‐7	
  Conclusion	
  ...........................................................................................................................	
  17	
  
I-­‐8	
  Experimental	
  .......................................................................................................................	
  18	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  I-­‐8-­‐1	
  Material	
  and	
  Method:	
  Protein	
  Expression	
  and	
  Purification	
  ...................................	
  18	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  I-­‐8-­‐2	
  	
  Kd	
  Determination	
  via	
  Fluorescence	
  Quenching	
  Assay	
  ...........................................	
  19	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  I-­‐8-­‐3	
  Crystallization	
  and	
  Structure	
  Determination	
  ..........................................................	
  20	
  
REFERENCES	
  .............................................................................................................................	
  24	
  
Chapter	
  II:	
  Domain	
  swapping	
  in	
  hCRBPII	
  ................................................................................	
  28	
  
II-­‐1	
  Introduction	
  .......................................................................................................................	
  28	
  
II-­‐2	
  Origin	
  of	
  dimerization	
  in	
  hCRBPII	
  .......................................................................................	
  35	
  
II-­‐3	
  Dimer	
  formation	
  of	
  other	
  hCRBPII	
  variants	
  .......................................................................	
  39	
  
II-­‐4	
  Structural	
  Studies	
  ...............................................................................................................	
  44	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐4-­‐1	
  Structural	
  analysis	
  reveals	
  an	
  extensive	
  domain	
  swapped	
  dimer	
  .........................	
  44	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐4-­‐2	
  Symmetry	
  VS	
  asymmetry	
  .......................................................................................	
  48	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐4-­‐3	
  Phase	
  Relationship	
  .................................................................................................	
  53	
  
II-­‐5	
  possible	
  mechanism	
  for	
  domain	
  swapping	
  in	
  hCRBPII	
  .......................................................	
  55	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐5-­‐1	
  Dimerization	
  and	
  structure	
  of	
  the	
  E72A	
  mutant	
  ...................................................	
  58	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐5-­‐2	
  Study	
  the	
  folding	
  pathway	
  .....................................................................................	
  60	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐5-­‐3	
  HCRBPII	
  folding	
  route	
  VS	
  other	
  iLBP	
  members.	
  .....................................................	
  63	
  
II-­‐6	
  Conformational	
  change	
  driven	
  by	
  ligand	
  binding	
  in	
  hCRBPII	
  .............................................	
  64	
  
II-­‐7	
  Experimental	
  ......................................................................................................................	
  73	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐1	
  Material	
  and	
  Method:	
  Site-­‐Directed	
  Mutagenesis	
  ................................................	
  73	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐2	
  Protein	
  Expression	
  and	
  Purification	
  of	
  DS	
  dimer	
  mutants	
  .....................................	
  76	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐3	
  Protein	
  Refolding	
  ...................................................................................................	
  77	
  
ix	
  

	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐4	
  Thermal	
  Melting	
  Curves	
  for	
  monomer	
  and	
  dimer	
  mutants.	
  ..................................	
  78	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐5	
  	
  Kd	
  determination	
  ...................................................................................................	
  80	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐6	
  Extinction	
  Coefficient	
  Determination	
  ....................................................................	
  80	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  II-­‐7-­‐7	
  Crystallization,	
  data	
  collection	
  and	
  refinement	
  .....................................................	
  80	
  
REFERENCES	
  .............................................................................................................................	
  85	
  

Chapter	
  III:	
  Can	
  we	
  predict	
  and	
  control	
  the	
  folding	
  product	
  of	
  proteins	
  through	
  their	
  amino	
  
acid	
  sequence	
  	
  (at	
  least	
  in	
  iLBP	
  family)?	
  .................................................................................	
  95	
  
III-­‐1	
  Introduction	
  ......................................................................................................................	
  95	
  
III-­‐2	
  Investigation	
  of	
  potential	
  domain	
  swapping	
  in	
  other	
  iLBP	
  family	
  members:	
  Case	
  Study	
  1,	
  
human	
  FABP4	
  ...........................................................................................................................	
  96	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐2-­‐1	
  History	
  of	
  FABP4	
  ...................................................................................................	
  96	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐2-­‐2	
  Investigation	
  on	
  hFABP4	
  bacterial	
  expression	
  ......................................................	
  97	
  
III-­‐3	
  Investigation	
  of	
  potential	
  domain	
  swapping	
  in	
  other	
  iLBP	
  family	
  members:	
  Case	
  Study	
  1,	
  
human	
  FABP5	
  ...........................................................................................................................	
  98	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐3-­‐1	
  History	
  of	
  hFABP5	
  .................................................................................................	
  98	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐3-­‐2	
  Investigate	
  domain	
  swapping	
  in	
  human	
  fatty	
  acid	
  binding	
  protein	
  5	
  .................	
  103	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐3-­‐3	
  Study	
  	
  domain	
  swapping	
  occurance	
  by	
  mutational	
  analysis	
  of	
  human	
  FABP5	
  ...	
  108	
  
III-­‐4	
  Experimental	
  ...................................................................................................................	
  111	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐4-­‐1	
  Material	
  and	
  Method:	
  Site-­‐Directed	
  Mutagenesis	
  .............................................	
  111	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐4-­‐2	
  Protein	
  Expression	
  and	
  Purification	
  of	
  hFABP4	
  and	
  hFABP5…………………………….113	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  III-­‐4-­‐3	
  Crystallization,	
  data	
  collection	
  and	
  refinement	
  ..................................................	
  114	
  
	
  	
  	
  	
  REFERENCES	
  ...........................................................................................................................	
  116	
  
	
  
Chapter	
  IV:	
  STRUCTURAL	
  INSIGHTS	
  INTO	
  EMISSION	
  REGULATION	
  OF	
  HCRBBPII-­‐BOUND	
  
SOLVATOCHROMIC	
  FLUOROPHORE	
  .....................................................................................	
  127	
  
IV-­‐1	
  Introduction	
  ....................................................................................................................	
  127	
  
IV-­‐2	
  Rational	
  behind	
  the	
  parent	
  template:	
  Q108K:K40L:T51V:T53S	
  ......................................	
  128	
  
IV-­‐3	
  Attempts	
  to	
  isolate	
  the	
  binding	
  cavity	
  ............................................................................	
  132	
  
IV-­‐4	
  Extensive	
  packing	
  of	
  the	
  ligand	
  and	
  removal	
  of	
  water	
  leads	
  to	
  bathochromic	
  shift	
  in	
  
absorption	
  and	
  emission.	
  .......................................................................................................	
  136	
  
IV-­‐5	
  Localization	
  of	
  charge	
  leads	
  to	
  a	
  blue	
  shift	
  in	
  absorption	
  and	
  emission	
  ........................	
  138	
  
IV-­‐6	
  Material	
  and	
  Methods	
  ....................................................................................................	
  141	
  
REFERENCES	
  ...........................................................................................................................	
  144	
  
	
  
Chapter	
  V:	
  Mutational	
  Studies	
  on	
  Starch	
  Branching	
  Enzyme	
  I	
  (RBEI)	
  to	
  understand	
  its	
  Surface	
  
Binding	
  Mechanism.	
  ............................................................................................................	
  157	
  
V-­‐1	
  introduction	
  .....................................................................................................................	
  157	
  
V-­‐2	
  Structure	
  of	
  the	
  both	
  N	
  and	
  C	
  terminus	
  truncated	
  RBEI:	
  ................................................	
  159	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐2-­‐1	
  Binding	
  Site	
  I:	
  Carbohydrate	
  Binding	
  Module	
  (CBM)	
  ...........................................	
  162	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐2-­‐2	
  New	
  Observed	
  Binding	
  Site	
  II	
  ..............................................................................	
  163	
  
V-­‐3	
  Mutational	
  studies	
  on	
  the	
  observed	
  binding	
  sites	
  and	
  active	
  site	
  of	
  RBEI	
  ......................	
  164	
  

x	
  

	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐3-­‐1	
  Effect	
  of	
  Active	
  Site	
  Mutation:	
  ............................................................................	
  165	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐3-­‐1	
  Effect	
  of	
  E534A	
  Mutation:	
  ...................................................................................	
  167	
  
V-­‐4	
  Experimental	
  ....................................................................................................................	
  170	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐4-­‐1	
  Material	
  and	
  Method	
  ..........................................................................................	
  170	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐4-­‐2	
  Activity	
  Assay	
  .......................................................................................................	
  173	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  V-­‐4-­‐3	
  Crystallization	
  of	
  RBEI	
  mutants	
  ............................................................................	
  175	
  
REFERENCES	
  ...........................................................................................................................	
  176	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

xi	
  

LIST	
  OF	
  TABLES	
  
	
  
Table	
  I-­‐1:	
  	
  	
  Data-­‐collection	
  and	
  refinement	
  statistics...................................................................22	
  
Table	
  II-­‐1:	
  	
  The	
  Effect	
  of	
  expression	
  temperature	
  and	
  in	
  vitro	
  refolding	
  protein	
  concentration	
  on	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  dimer/monomer	
  	
  (D/M)	
  ratio...............................................................................40	
  
	
  
Table	
  II-­‐2:	
  Monomer/dimer	
  ratio	
  of	
  various	
  residue	
  60	
  mutants.	
  During	
  bacterial	
  expression	
  at	
  	
  
	
  	
  	
  	
  	
  	
  30	
  °C............................................................................................................................42	
  
	
  
Table	
  II-­‐3:	
  	
  	
  Summary	
  of	
  DS	
  dimer	
  crystal	
  structures	
  of	
  all	
  hCRBPII	
  mutants...............................66	
  
	
  
Table	
  II-­‐4	
  :	
  	
  PCR	
  protocol	
  for	
  hCRABPII	
  and	
  hCRBPII	
  mutagenesis...............................................74	
  
	
  
Table	
  II-­‐5:	
  	
  Anion	
  Exchange	
  purification	
  protocol	
  for	
  hCRBPII.	
  The	
  buffer	
  used	
  for	
  above	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  protocol	
  is	
  50	
  mM	
  Tris,	
  pH	
  is	
  adjusted	
  automatically.	
  The	
  proteins	
  elute	
  between	
  4%	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  -­‐8	
  %	
  2M	
  NaCl.	
  .............................................................................................................77	
  
	
  
Table	
  II-­‐6:	
  Calculated	
  percentage	
  of	
  helix/strand	
  for	
  each	
  denaturant	
  (GD.HCl)	
  concentration	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  during	
  unfolding	
  experiment.	
  220nm	
  corresponds	
  to	
  beta	
  sheets.	
  222nm/208nm	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  corresponds	
  to	
  α-­‐helixes.............................................................................................80	
  
	
  
Table	
  II-­‐7:	
  	
  X-­‐ray	
  crystallographic	
  data	
  and	
  refinement	
  statistics	
  for	
  hCRBPII	
  	
  mutants..............82	
  
	
  

Table	
  III-­‐1:	
  	
  Crystallographic	
  data	
  of	
  hFABP5	
  (pseudo	
  monomer).............................................115	
  
	
  
Table	
  IV-­‐1:	
  Spectroscopic	
  change	
  of	
  hCRBPII	
  variants	
  bound	
  ThioFluor....................................132	
  
	
  
Table	
  IV-­‐2:	
  	
  X-­‐ray	
  crystallographic	
  data	
  and	
  refinement	
  statistics	
  for	
  monomeric	
  hCRBPII	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  mutants....................................................................................................................143	
  
	
  
Table	
  V-­‐1:	
  	
  Residues	
  involved	
  in	
  the	
  interactions	
  in	
  Binding	
  sites	
  I,	
  II	
  and	
  III	
  of	
  RBEI-­‐	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  M12........................................................................................................................165	
   	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
Table	
  V-­‐2:	
  	
  More	
  suggested	
  mutations	
  to	
  study	
  for	
  residues	
  interacting	
  with	
  M12	
  in	
  binding	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  site	
  II	
  of	
  RBEI.	
  ............................................................................................................170	
  
	
  
	
  
	
  

xii	
  

LIST	
  OF	
  FIGURES	
  
	
  
	
  
Figure	
  I-­‐1:	
  	
  a.	
  An	
  overlay	
  the	
  structures	
  of	
  hCRBPII	
  (green	
  and	
  cyan),	
  hCRBPIII	
  (1GGL,	
  salmon),	
  	
  	
  
	
  	
  	
  	
  	
  	
  hCRBPIV	
  (1LPJ,	
  yellow)	
  and	
  hCRBPI	
  (1KGL,	
  pink)	
  b.	
  An	
  overlay	
  of	
  the	
  structures	
  of	
  apo	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  hCRBPII	
  (2RCQ,	
  gray),	
  wt	
  hCRBPII	
  bound	
  to	
  all-­‐trans-­‐retinal	
  (4QYP,	
  green)	
  and	
  to	
  all-­‐	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  trans-­‐retinol	
  (4QYN,cyan)………………………………………………………………………………………….2	
  
	
  
Figure	
  I-­‐2:	
  The	
  multiple	
  sequence	
  alignment	
  between	
  human,	
  rat	
  CRBPs	
  and	
  zebra	
  fish	
  CRBP...3	
  
	
  
Figure	
  I-­‐3:	
  Ligand	
  binding	
  measurements	
  using	
  the	
  fluorescence	
  quenching	
  of	
  tryptophan	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
(excitation	
   at	
   280	
   nm,	
   emission	
   at	
   350	
   nm).	
   a.	
  Titration	
   of	
   WT-­‐hCRBPII	
   with	
   all-­‐trans-­‐	
  	
  	
  	
  
retinol.	
   b.	
   Titration	
   of	
   hCRBPII	
   T51I	
   mutant	
   with	
   all-­‐trans-­‐retinol.	
   c.	
   Titration	
   of	
  
hCRBPII	
  T51I	
  mutant	
  with	
  all-­‐trans-­‐retinal.	
  	
  d.	
  Titration	
  of	
  WT-­‐hCRBPII	
  with	
  all-­‐trans-­‐
retinal.	
  ………………………………………………………………………………………………………………………5	
  
	
  
Figure	
  I-­‐4:	
  	
  Retinal	
  binding	
  in	
  hCRBPII.	
  a.	
  An	
  overlay	
  of	
  retinal	
  from	
  mol	
  A	
  (cyan),	
  mol	
  B	
  (blue)	
  	
  	
  	
  
and	
  Mol	
  C	
  (pink),	
  showing	
  the	
  similarity	
  in	
  binding	
  in	
  the	
  three.	
  b.	
  Stick	
  	
  	
  	
  
representation	
  of	
  the	
  interaction	
  between	
  	
  Phe16,	
  Leu77	
  (both	
  with	
  green	
  carbons,	
  N	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  blue	
  and	
  O,	
  red)	
  and	
  the	
  retinal	
  ionone	
  ring	
  (blue	
  carbons,	
  Mol	
  B).	
  c.	
  Space	
  filling	
  	
  	
  
	
  	
  	
  	
  	
  depiction	
  of	
  b.	
  ………………………………………………………………………………………………………..…7	
  
	
  
Figure	
  I-­‐5:	
  	
  The	
  water	
  mediated	
  interaction	
  of	
  the	
  retinal	
  hydroxyl	
  group	
  in	
  hCRBPII.	
  	
  Atoms	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
colored	
  by	
  type,	
  hydrogen	
  bonded	
  distances	
  are	
  shown.………………………………………….8	
  
	
  
Figure	
  I-­‐6:	
  a.	
  The	
  contoured	
  2Fo-­‐Fc	
  electron	
  density	
  map	
  at	
  1.0	
  σ	
  for	
  rearranged	
  retinol	
  from	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  data	
  collected	
  at	
  11	
  KeV/the	
  highest	
  flux	
  (4QYN,	
  blue).	
  b.	
  The	
  chemical	
  structure	
  of	
  a	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  possible	
  retinoid	
  derivative	
  consistent	
  with	
  the	
  crystallographic	
  data.	
  c.	
  The	
  overlay	
  of	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  bonafide	
  retinol	
  structure	
  (chain	
  C)	
  from	
  data	
  collected	
  with	
  an	
  attenuated	
  X-­‐ray	
  	
  
	
  	
  	
  	
  	
  beam	
  at	
  an	
  energy	
  of	
  7	
  KeV	
  (4QZT,	
  Pink)	
  and	
  the	
  structure	
  of	
  rearraged	
  retinoid	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  derivative	
  (chain	
  B)	
  from	
  data	
  collected	
  at	
  11	
  KeV	
  and	
  the	
  highest	
  flux	
  (4QYN,	
  blue).	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  d.	
  The	
  structures	
  of	
  the	
  noncanonical	
  retinoid	
  (refined	
  in	
  two	
  conformations)	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  previously	
  seen	
  in	
  the	
  human	
  CRBPII	
  (2RCT,	
  yellow),	
  and	
  our	
  current	
  rearranged	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structure	
  (4QYN,	
  blue)	
  were	
  overlaid.	
  e.	
  The	
  structures	
  of	
  the	
  canonical	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  noncanonical	
  retinoid	
  previously	
  seen	
  in	
  the	
  zebrafish	
  CRBP	
  (1KQW,	
  green),	
  and	
  our	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  current	
  rearranged	
  structure	
  (4QYN,	
  blue)	
  were	
  overlaid.	
  f.	
  The	
  overlaid	
  structures	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  canonical	
  retinol	
  obtained	
  in	
  the	
  rat	
  CRBPII	
  (1OPB,	
  purple)	
  and	
  our	
  current	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  bonafide	
  all-­‐trans-­‐retinol	
  structure	
  in	
  human	
  CRBPII	
  (4QZT,	
  pink).…………………………….9	
  
	
  
Figure	
  I-­‐7:	
  	
  a.	
  The	
  rearranged	
  retinol	
  from	
  data	
  collected	
  at	
  11	
  KeV	
  and	
  the	
  highest	
  flux	
  (4QYN,	
  	
  
	
  	
  	
  	
  	
  	
  blue)	
  and	
  the	
  interaction	
  of	
  alcohol	
  moiety	
  with	
  the	
  neighboring	
  residues	
  of	
  Gln108	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  Lys40.	
  b.	
  The	
  structures	
  of	
  the	
  retinal	
  bound	
  hCRBPII	
  (4QYP,	
  cyan,	
  chain	
  B),	
  and	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  

xiii	
  

our	
   current	
   rearranged	
   retinol	
   structure	
   (4QYN,	
   blue)	
   were	
   overlaid.	
   The	
   order	
   water	
  
molecule	
   (W)	
   seen	
   in	
   all-­‐trans-­‐retinal	
   crystal	
   structure,	
   is	
   occupied	
   by	
   the	
   hydroxyl	
  
group	
   of	
   the	
   noncanonical	
   retinoid	
   (indicated	
   in	
   dashed	
   circle).	
   Similar	
   case	
   in	
   retinol	
  
structure	
   from	
   data	
   collected	
   at	
   7	
   KeV	
   and	
   an	
   attenuated	
   beam	
   has	
   been	
   seen.	
   c.	
  
Interactions	
  around	
  Gln108	
  of	
  retinol	
  bound	
  rat	
  CRBPII	
  (1OPB,	
  purple)..………………...10	
  

	
  
Figure	
  I-­‐8:	
  	
  This	
  experiment	
  performed	
  by	
  Dr.	
  Yapici	
  a.	
  UV-­‐vis	
  spectrum	
  of	
  all-­‐trans-­‐retinol.	
  	
  b.	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  The	
  HPLC	
  trace	
  of	
  all-­‐trans-­‐retinol.	
  c.	
  	
  UV-­‐vis	
  spectrum	
  of	
  the	
  retinol	
  extracted	
  from	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  crystals	
  and	
  purified	
  by	
  HPLC.	
  d.	
  HPLC	
  chromatogram	
  of	
  the	
  retinol	
  extracted	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  from	
  crystals.	
  e.	
  	
  HPLC	
  chromatogram	
  of	
  bonafide	
  all-­‐trans-­‐retinol	
  and	
  the	
  retinol	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  extracted	
  from	
  crystals.	
  Showing	
  that	
  they	
  are	
  the	
  same	
  species.............................12	
  
	
  
Figure	
  I-­‐9:	
  A	
  comparison	
  of	
  the	
  retinol	
  and	
  retinoid	
  moieties	
  found	
  in	
  the	
  crystal	
  structures	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  human,	
  zebra	
  fish	
  and	
  rat	
  CRBPII.	
  a.	
  The	
  retinol	
  (left)	
  and	
  retinoid	
  (right)	
  moieties	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  found	
  in	
  the	
  previously	
  published	
  hCRBPII	
  (2RCT,	
  yellow).	
  The	
  value	
  of	
  128°	
  torsion	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  angle	
  (ψ)	
  about	
  the	
  C13-­‐C14	
  bond	
  is	
  inconsistent	
  with	
  a	
  double	
  bond	
  (180°).	
  b.	
  	
  The	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  retinol	
  and	
  retinoid	
  moieties	
  seen	
  in	
  the	
  crystal	
  structure	
  of	
  zebrafish	
  CRBP	
  (1KQW,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  green).	
  	
  The	
  torsion	
  angle	
  about	
  the	
  C13-­‐C14	
  bond	
  is	
  also	
  inconsistent	
  with	
  a	
  double	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  bond.	
  c.	
  The	
  retinol	
  structure	
  in	
  the	
  rat	
  CRBPII	
  (1OPB,	
  purple).	
  d.	
  	
  The	
  retinol	
  found	
  in	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  X-­‐ray	
  structure	
  of	
  hCRBPII	
  using	
  data	
  collected	
  at	
  an	
  energy	
  of	
  7	
  KeV	
  and	
  an	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  attenuated	
  beam	
  (4QZT,	
  pink)	
  with	
  an	
  approximate	
  180°	
  torsion	
  angle	
  around	
  C13-­‐	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  C14	
  bond.	
  …………………………….......................................................................................14	
  
	
  
Figure	
  I-­‐10:	
  Retinal	
  versus	
  retinol	
  binding	
  in	
  hCRBPII.	
  a.	
  The	
  structure	
  of	
  the	
  retinol-­‐bound	
  WT	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hCRBPII.	
  b.	
  The	
  2Fo-­‐Fc	
  electron	
  density	
  map	
  contoured	
  at	
  1.0σ	
  around	
  retinol	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  7Kev	
  data.	
  c.	
  The	
  structure	
  of	
  all-­‐trans-­‐retinal	
  bound	
  to	
  WT	
  hCRBPII.	
  d.	
  The	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structures	
  of	
  all-­‐trans-­‐retinal	
  (green)	
  and	
  retinol	
  (magenta).	
  The	
  position	
  of	
  an	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  ordered	
  water	
  molecule	
  (W)	
  occupied	
  by	
  the	
  hydroxyl	
  group	
  of	
  retinol	
  is	
  indicated	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  in	
  dashed	
  circle.........................................................................................................15	
  
	
  
Figure	
  I-­‐11:	
  The	
  structures	
  of	
  retinol-­‐bound	
  rat	
  CRBPI	
  (magenta	
  carbons	
  and	
  hydrogen	
  bonds)	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  retinal-­‐bound	
  hCRBPII	
  (green	
  carbons	
  and	
  hydrogen-­‐bonds)	
  are	
  overlayed.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Amino	
  acid	
  labels	
  are	
  colored	
  similarly	
  when	
  different	
  in	
  the	
  two	
  structures.	
  	
  Note	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  change	
  in	
  the	
  water	
  network	
  introduced	
  by	
  Phe3…...........................................16	
  
	
  
Figure	
  II-­‐1:	
  Cartoon	
  illustration	
  of	
  3D	
  domain	
  swapping.	
  Identical	
  structural	
  units	
  swap	
  between	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  2	
  monomers,	
  followed	
  by	
  creating	
  the	
  open	
  interface	
  between	
  them......................28	
  
	
  
Figure	
  II-­‐2:	
  a.	
  Left:	
  DT	
  monomer	
  (PDB	
  ID	
  4OW6).	
  Right:	
  DT	
  dimer	
  (4AE1).	
  The	
  two	
  subunits	
  are	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  blue	
  and	
  pink.	
  In	
  acidic	
  pH	
  equilibrium	
  goes	
  toward	
  DS	
  dimer	
  formation.	
  b.	
  Left	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structure	
  of	
  DS	
  dimer	
  of	
  RnaseA	
  (PDB	
  1A2W)	
  and	
  cyclic	
  DS	
  trimes	
  (PDB	
  5SRA).........30	
  
	
  
Figure	
  II-­‐3:	
  	
  The	
  hCRBPII	
  DS	
  dimer................................................................................................34	
  
	
  
xiv	
  

Figure	
  II-­‐4:	
  Trajectory	
  of	
  Tyr	
  60,	
  Glu	
  72	
  (both	
  shown	
  as	
  stick	
  representations,	
  colored	
  blue)	
  in	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinal...............................................................................36	
  
	
  
Figure	
  II-­‐5:	
  a.	
  IEX	
  chromatogram	
  of	
  the	
  KLY60W,	
  monitored	
  at	
  280nm.	
  b.	
  The	
  UV-­‐Vis	
  spectra	
  of	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  80mM	
  salt	
  elution	
  (red)	
  and	
  the	
  160mM	
  salt	
  elution	
  (blue)	
  of	
  KLY60W	
  incubated	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  all-­‐trans-­‐retinal.	
  c.	
  The	
  UV-­‐Vis	
  spectra	
  of	
  the	
  refolded	
  80mM	
  eluent	
  (low	
  salt,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  blue)	
  after	
  denaturation	
  and	
  refolding	
  at	
  0.03	
  mM	
  concentration	
  and	
  the	
  native	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  160mM	
  (high	
  salt,	
  green)	
  and	
  native	
  80mM	
  elution	
  (red),	
  all	
  bound	
  to	
  all-­‐trans-­‐	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  retinal.	
  d.	
  The	
  size	
  exclusion	
  chromatogram	
  (Superdex	
  200	
  16/75	
  column)	
  for	
  dimer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  (the	
  160mM	
  elution)	
  KLY60W,	
  after	
  maintaining	
  the	
  protein	
  at	
  room	
  temperature	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  for	
  several	
  days,	
  showing	
  only	
  the	
  monomer	
  size	
  peak.	
  e.	
  Gel	
  filtration	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  chromatogram	
  of	
  monomer	
  (the	
  80mM	
  elution)	
  of	
  KLY60W,	
  showing	
  only	
  the	
  	
  dimer	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  fractions.	
  f-­‐i.	
  Ion	
  exchange	
  chromatographs	
  of	
  WT	
  and	
  various	
  hCRBPII	
  mutants,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monitored	
  at	
  280	
  nm.	
  f.	
  Y60I	
  at	
  30°C.	
  g.	
  Y60L	
  at	
  30°C.	
  h.	
  Y60L	
  at	
  25°C.	
  i.	
  WT	
  hCRBPII	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  expressed	
  at	
  30	
  °C......................................................................................................37	
  
	
  
Figure	
  II-­‐6:	
  	
  Thermal	
  melting	
  curves:	
  a.	
  WT	
  hCRBPII	
  dimer	
  (θ	
  measured	
  at	
  220	
  nm),	
  Tm	
  =	
  52	
  °C;	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  b.	
  Y60L	
  	
  dimer	
  (θ	
  measured	
  at	
  227	
  nm),	
  Tm	
  =	
  56	
  °C;	
  c.	
  KLY60W	
  dimer	
  (θ	
  measured	
  at	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  220	
  nm),	
  Tm	
  =	
  69°C.	
  d.	
  WT	
  hCRBPII	
  monomer	
  (θ	
  measured	
  at	
  220	
  nm);	
  e.	
  Y60L	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monomer	
  (θ	
  measured	
  at	
  215nm);	
  f.	
  KLY60W	
  monomer	
  (θ	
  measured	
  at	
  220	
  nm).	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Note	
  that	
  most	
  of	
  the	
  secondary	
  structure	
  is	
  preserved	
  in	
  both	
  the	
  WT	
  and	
  KLY60W	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monomers,	
  even	
  at	
  the	
  highest	
  temperature	
  achieved	
  (CD	
  intensities	
  of-­‐65mdeg	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  -­‐50mdeg	
  respectively).	
  Thus	
  melting	
  curves	
  can	
  only	
  give	
  a	
  lower	
  limit	
  for	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  TM,	
  consistent	
  with	
  the	
  view	
  that	
  the	
  monomers	
  are	
  more	
  stable	
  than	
  the	
  dimers.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  The	
  apparent	
  lack	
  of	
  a	
  Tm	
  for	
  the	
  monomers	
  might	
  suggest	
  the	
  presence	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  thermally	
  induced	
  folding	
  intermediates.	
  Clearly	
  the	
  unfolding	
  process	
  is	
  not	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monolithic	
  as	
  that	
  observed	
  for	
  the	
  dimeric	
  species.	
  This	
  lends	
  further	
  support	
  to	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  suggestion	
  that	
  the	
  open	
  monomer,	
  which	
  is	
  presumably	
  obtained	
  from	
  melting	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  of	
  the	
  dimer,	
  has	
  a	
  different	
  folding	
  trajectory	
  (and	
  consequently	
  a	
  different	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  unfolding	
  trajectory)	
  as	
  compared	
  to	
  the	
  monomeric	
  species..................................39	
  
	
  
Figure	
  II-­‐7:	
  	
  Size	
  exclusion	
  chromatography	
  after	
  protein	
  denaturation	
  and	
  refolding,	
  all	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monitored	
  at	
  280nm.	
  	
  	
  a.	
  Y60L	
  mutant	
  refolded	
  at	
  0.03	
  mM	
  concentration.	
  b.	
  Y60L	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  refolded	
  at	
  0.13	
  mM	
  concentration.	
  c.	
  Y60L	
  refolded	
  at	
  0.4	
  mM	
  concentration.	
  d.	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  WT	
  hCRBPII	
  refolded	
  at	
  0.03	
  mM	
  concentration.	
  e.	
  WT	
  hCRBPII	
  refolded	
  at	
  0.13	
  mM	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  concentration.	
  f.	
  WT	
  hCRBPII	
  refolded	
  at	
  0.3	
  mM	
  concentration.	
  ...........................41	
  
	
  
Figure	
  II-­‐8:	
  	
  a.The	
  size	
  exclusion	
  chromatogram	
  for	
  dimer	
  of	
  WT	
  hCRBPII,	
  after	
  maintaining	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  protein	
  at	
  room	
  temperature	
  for	
  several	
  days,	
  showing	
  only	
  the	
  dimer	
  species.	
  b.	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Gel	
  filtration	
  chromatogram	
  of	
  monomer	
  of	
  WT,	
  showing	
  only	
  the	
  monomer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  protein.......................................................................................................................43	
  
	
  
Figure	
  II-­‐9:	
  	
  Ligand	
  binding,	
  monitored	
  by	
  Tryptophan	
  fluorescence	
  quenching	
  of	
  monomeric	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  dimeric	
  hCRBPII.	
  a.	
  Monomer	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinol.	
  b.	
  Dimer	
  WT	
  	
  
xv	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hCRBPII	
  with	
  all-­‐trans-­‐retinol.	
  c.	
  Monomer	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinal.	
  d.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Dimer	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinal....................................................................44	
  
	
  
Figure	
  II-­‐10:	
  Crystal	
  structure	
  of	
  dimer	
  KLY60W	
  mutant	
  and	
  its	
  monmer	
  components	
  a.	
  Chain	
  A	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  of	
  KLY60W	
  (where	
  only	
  the	
  Thr56	
  psi	
  angle	
  differs	
  from	
  the	
  closed	
  monomer	
  form)	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  the	
  electron	
  density,	
  contoured	
  at	
  1.0	
  σ,	
  of	
  the	
  hinge	
  loop	
  region	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  encompassing	
  amino	
  acids	
  55-­‐60.	
  	
  Atoms	
  colored	
  by	
  type,	
  N,	
  blue,	
  O,	
  red,	
  C,	
  cyan.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  b.	
  Chain	
  B,	
  showing	
  the	
  electron	
  density,	
  contoured	
  at	
  1.0	
  σ,	
  of	
  the	
  hinge	
  loop	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  region	
  encompassing	
  amino	
  acids	
  55-­‐60.	
  Atoms	
  colored	
  	
  	
  	
  	
  	
  by	
  	
  type,	
  N,	
  blue,	
  O,	
  red,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  C,	
  cyan.	
  c.	
  The	
  complete	
  structure	
  of	
  the	
  KLY60W	
  dimer:	
  In	
  chain	
  A	
  	
  (cyan)	
  Trp60	
  is	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  pointing	
  inside	
  the	
  binding	
  pocket,	
  while	
  in	
  chain	
  B	
  (pink)	
  Trp60	
  is	
  pointing	
  toward	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  solvent...............................................................................................................45	
  
	
  
Figure	
  II-­‐11:	
  	
  An	
  overlay	
  of	
  the	
  dimeric	
  (cyan)	
  and	
  monomeric	
  (red)	
  KLY60W	
  mutant	
  structures.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Res56	
  for	
  both	
  structures	
  is	
  shown	
  in	
  sticks.	
  The	
  hinge	
  loop	
  region	
  in	
  the	
  monomer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  dimer	
  is	
  encircled,	
  showing	
  the	
  dramatic	
  difference	
  in	
  the	
  two	
  structures......46	
  
	
  
Figure	
  II-­‐12:	
  	
  a-­‐e:	
  Torsion	
  angle	
  differences	
  between	
  monomer	
  and	
  dimer	
  differences	
  of	
  Phi	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  angle	
  are	
  in	
  blue	
  and	
  psi	
  angle	
  differences	
  are	
  in	
  red	
  in	
  each	
  case.	
  	
  a.	
  Thr56.	
  b.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Phe57.	
  c.	
  Arg58.	
  d.	
  Asn59.	
  e.	
  Res60………………………………………………………………..……47	
  
	
  
Figure	
  II-­‐13:	
  The	
  differences	
  in	
  the	
  asymmetric	
  and	
  symmetric	
  structures	
  of	
  hCRBPII	
  dimers	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Res60	
  in	
  all	
  structures	
  is	
  shown	
  in	
  sticks	
  a.	
  The	
  asymmetric	
  structures:	
  The	
  overlay	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structure	
  of	
  domain	
  swapped	
  dimer	
  KLY60W	
  (cyan)	
  and	
  Y60W	
  (pink).	
  In	
  chain	
  A	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer	
  KLY60W,	
  Trp60	
  is	
  pointing	
  inside	
  the	
  binding	
  pocket.	
  b.	
  The	
  symmetric	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structures.	
  The	
  overlay	
  of	
  WT	
  hCRBPII	
  (purple)	
  and	
  Y60L	
  (yellow)	
  domain	
  swapped	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer.	
  c.	
  Comparing	
  asymmetric	
  (dimer	
  KLY60W	
  in	
  cyan)	
  and	
  symmetric	
  dimer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  (dimer	
  Y60L	
  in	
  yellow)	
  by	
  looking	
  down	
  the	
  two-­‐fold	
  axis	
  and	
  it	
  shows	
  the	
  different	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  position	
  of	
  helices	
  in	
  N-­‐terminus.	
  d.	
  One	
  chain	
  of	
  dimer	
  Y60L	
  (yellow)	
  and	
  both	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  chains	
  of	
  asymmetric	
  dimer	
  Y60W	
  (chain	
  A	
  in	
  cyan	
  and	
  chain	
  B	
  in	
  red)	
  are	
  overlaid	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  at	
  their	
  C-­‐terminus...................................................................................................48	
  
	
  
Figure	
  II-­‐14:	
  	
  Torsion	
  angle	
  deviations	
  outside	
  the	
  hinge-­‐loop	
  region	
  define	
  the	
  relative	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  orientation	
  of	
  the	
  two	
  domains	
  of	
  the	
  dimer,	
  with	
  the	
  highest	
  deviation	
  seen	
  in	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Asn59.	
  a.	
  	
  The	
  WT	
  (purple)	
  and	
  Y60L	
  (yellow)	
  dimers	
  are	
  overlaid.	
  	
  Inset,	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hydrogen	
  bond	
  between	
  Tyr60	
  and	
  Glu72	
  compared	
  to	
  the	
  “flipped	
  out”	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  conformation	
  of	
  Leu60.	
  b.	
  The	
  critical	
  Tyr60	
  and	
  Asn59	
  region	
  in	
  the	
  WT	
  hCRBPII	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer,	
  showing	
  the	
  key	
  phi/psi	
  angles.	
  c.	
  The	
  same	
  region	
  in	
  the	
  Y60L	
  dimer,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  showing	
  the	
  “flipped	
  out”	
  Leu60	
  and	
  the	
  key	
  phi/psi	
  angles.	
  	
  Comparison	
  of	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  two	
  shows	
  how	
  the	
  large	
  difference	
  in	
  the	
  phi	
  angle	
  is	
  compensated	
  for	
  in	
  the	
  WT	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  N59	
  psi	
  angle,	
  keeping	
  the	
  main	
  chain	
  of	
  the	
  two	
  on	
  a	
  similar	
  trajectory……………50	
  
	
  
	
  Figure	
  II-­‐15:	
  Residues	
  56-­‐63	
  in	
  the	
  chains	
  A	
  (green,	
  top)	
  and	
  B	
  (magenta,	
  bottom)	
  of	
  Y60W	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hCRBPII.....................................................................................................................51	
  
xvi	
  

	
  
Figure	
  II-­‐16:	
  Overview	
  of	
  dimer	
  formation	
  in	
  asymmetric	
  structures:	
  Two	
  chains	
  of	
  Y60W	
  dimer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  are	
  labeled	
  in	
  the	
  picture.	
  The	
  same	
  chain	
  (molecule	
  A	
  and/or	
  B)	
  in	
  this	
  asymmetric	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimers	
  cannot	
  form	
  dimers	
  since	
  they	
  clash.	
  However,	
  the	
  interaction	
  of	
  different	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  chain	
  with	
  each	
  other	
  (molecule	
  A	
  and	
  B)	
  could	
  lead	
  to	
  proper	
  DS	
  dimer	
  formation	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  the	
  different	
  orientation....................................................................................53	
  
	
  
Figure	
  II-­‐17:	
  Schematic	
  representation	
  of	
  “phase	
  relationship”	
  in	
  DS	
  dimerization.	
  Top:	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monomer,	
  bottom:	
  dimer	
  with	
  residue	
  60	
  as	
  mutant..............................................54	
  
	
  
Figure	
  II-­‐18:	
  	
  DS	
  dimerization	
  requires:	
  1.	
  Rotation	
  about	
  Thr56	
  psi.	
  2.	
  Orientation	
  of	
  the	
  two	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  halves	
  to	
  accommodate	
  dimerization.	
  	
  3.	
  Rephasing	
  of	
  the	
  connecting	
  strand.......56	
  
	
  
Figure	
  II-­‐19:	
  	
  DS	
  dimer	
  vs.	
  monomer	
  formation	
  in	
  WT	
  hCRBP	
  II..................................................57	
  
	
  
Figure	
  II-­‐20:	
  	
  Size	
  exclusion	
  chromatography	
  of	
  low	
  salt	
  and	
  high	
  salt	
  (80	
  mM	
  and	
  160	
  mM	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  NaCl,)	
  E72A	
  hCRBPII	
  respectively,	
  expressed	
  at	
  25°C	
  (since	
  30°C	
  expressed	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  produced	
  virtually	
  no	
  soluble	
  protein).....................................................................58	
  
	
  
Figure	
  II-­‐21:	
  a.	
  Overlay	
  of	
  chain	
  A	
  of	
  the	
  E72A	
  dimer	
  (green)	
  and	
  the	
  Y60L	
  dimer	
  (yellow).	
  Note	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  that	
  both	
  show	
  residue	
  60	
  pointing	
  toward	
  the	
  solvent.	
  b.	
  Overlay	
  of	
  the	
  E72A	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  chain	
  A	
  (blue)	
  and	
  WT	
  (green)	
  dimers.	
  Inset	
  shows	
  that	
  without	
  the	
  Glu	
  72	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hydrogen	
  bond	
  donor,	
  residue	
  60	
  is	
  free	
  to	
  flip	
  out,	
  leading	
  to	
  a	
  more	
  relaxed	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer	
  structure.	
  c.	
  Two	
  chains	
  of	
  the	
  E72A	
  dimer	
  (green	
  and	
  grey)	
  are	
  overlaid,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  showing	
  their	
  strong	
  similarity.................................................................................60	
  
	
  
Figure	
  II-­‐22:	
  Equilibrium	
  Unfolding	
  Experiment	
  for	
  wild-­‐type	
  CRBPII	
  in	
  presence	
  of	
  different	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  concentration	
  of	
  Gd-­‐HCl,	
  a.-­‐b.	
  monitoring	
  by	
  Circular	
  Dichroism	
  (CD).	
  c-­‐d.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Tryptophan	
  fluorescence	
  spectroscopy,	
  right,	
  performed	
  with	
  WT	
  hCRBPII,	
  to	
  study	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  presence	
  an	
  intermediate	
  in	
  the	
  folding	
  pathway.	
  	
  Purple	
  solid	
  line	
  represents	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  the	
  polynomial	
  fit	
  (3rd	
  degree)	
  in	
  b	
  and	
  d.	
  Both	
  b	
  and	
  d	
  spectra	
  are	
  in	
  correlate	
  with	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  previous	
  studies.	
  (1)	
  	
  .................................................................................................62	
  
	
  
Figure	
  II-­‐23:	
  	
  overview	
  picture	
  of	
  hCRABPI	
  folding	
  pathway	
  compare	
  to	
  hCRBPII.......................64	
  
	
  
Figure	
  II-­‐24:	
  	
  A	
  comparison	
  of	
  four	
  symmetric	
  dimers,	
  Y60L	
  (orange),	
  WT	
  (red),	
  E72A	
  (cyan)	
  and	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Q108K,K40D	
  (green)...............................................................................................,..66	
  
	
  
Figure	
  II-­‐25:	
  Overlay	
  of	
  holo	
  Q108K:K40D	
  (chains	
  F	
  and	
  A,	
  both	
  shown	
  	
  in	
  orange)	
  and	
  apo	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Q108K:T51D	
  (shown	
  in	
  blue)	
  showing	
  the	
  large	
  motion	
  of	
  helix	
  1	
  upon	
  ligand	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  binding	
  (about	
  15	
  Å).	
  Bound	
  retinal	
  molecules	
  in	
  holo	
  structures	
  are	
  shown	
  as	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  transparent	
  spherical	
  models.	
  Note,	
  complete	
  dimer	
  in	
  the	
  apo	
  structures	
  was	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  generated	
  by	
  crystallographic	
  two-­‐fold	
  symmetry	
  operation.	
  ................................67	
  
	
  
xvii	
  

Figure	
  II-­‐26:	
  	
  Ligand	
  binding	
  leads	
  Asn59	
  to	
  flip	
  out	
  of	
  the	
  binding	
  pocket	
  due	
  to	
  steric	
  effect.	
  a.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Overlay	
  of	
  Q108K:T51D-­‐apo	
  (green,	
  obtained	
  by	
  ALireza)	
  and	
  Q108K:K40D-­‐holo	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  (orange).	
  Allosteric	
  conformational	
  change	
  between	
  apo	
  and	
  holo	
  dimers	
  is	
  caused	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  by	
  the	
  orientation	
  of	
  the	
  Tyr	
  60	
  and	
  Asn	
  59	
  sidechains.	
  Asn59	
  would	
  sterically	
  clash	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  the	
  bound	
  retinylidene,	
  which	
  leads	
  to	
  the	
  flipped-­‐out	
  conformation	
  of	
  Asn59	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  flipped-­‐in	
  conformation	
  of	
  Tyr60.	
  b.	
  Overlay	
  of	
  WT	
  apo	
  monomer	
  (pink,	
  PDB	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  code:	
  2RCQ)	
  and	
  WT	
  apo	
  hCRBPII	
  dimer	
  (green,	
  PDB	
  code:	
  4ZH9).	
  The	
  “flipped	
  in”	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Asn59	
  conformation	
  is	
  required	
  to	
  adopt	
  the	
  orientation	
  to	
  form	
  the	
  symmetric	
  DS	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer.	
  c.	
  The	
  similar	
  trajectory	
  of	
  retinal	
  in	
  monomer	
  and	
  dimer	
  structures	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hCRBPII	
  mutants.	
  Q108K:K40L	
  monomer	
  (purple,	
  PDB	
  code:	
  4RUU)	
  Q108K:K40D	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer	
  (orange)	
  and	
  Q108K:K40L:T51DD	
  dimer	
  (pink,	
  obtained	
  by	
  Dr.Nosrati)	
  which	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  are	
  all	
  bound	
  to	
  retinal	
  via	
  a	
  covalent	
  bond	
  to	
  Lys108.	
  d.	
  Overlay	
  of	
  WT	
  monomer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hCRBPII	
  bound	
  to	
  retinol	
  (green,	
  PDB	
  code:	
  4QZT)	
  with	
  retinal-­‐bound	
  Q108K:K40D	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimer	
  (orange)........................................................................................................68	
  
	
  
Figure	
  II-­‐27:	
  Overlay	
  of	
  WT-­‐hCRBPII	
  (holo)	
  DS	
  dimer	
  shown	
  by	
  dark	
  blue	
  and	
  WT-­‐hCRBPII	
  (apo)	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  (PDB	
  code:4ZH9)	
  shown	
  by	
  green............................................................................69	
  
	
  
Figure	
  II-­‐28:	
  Overlay	
  of	
  different	
  chains	
  in	
  holo	
  Q108K:K40D	
  DS	
  dimer.	
  a.	
  Chains	
  A	
  shown	
  by	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  orange	
  and	
  I	
  shown	
  by	
  green.	
  b.	
  Chains	
  A	
  shown	
  by	
  orange	
  and	
  L	
  shown	
  by	
  blue.	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  c.	
  ChainA	
  shown	
  by	
  orange	
  and	
  D	
  shown	
  by	
  pale	
  purple	
  d.	
  Chains	
  A	
  shown	
  by	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  orange	
  	
  and	
  G	
  shown	
  by	
  purple.	
  e.	
  Overlay	
  of	
  chains	
  A	
  shown	
  by	
  orange	
  and	
  B	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  shown	
  by	
  yellow.	
  f.	
  Overlay	
  of	
  chains	
  A	
  shown	
  by	
  green	
  and	
  G	
  shown	
  by	
  cyan.	
  g-­‐l:	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  E72…Y60	
  Hydrogen	
  boning	
  motifs	
  in	
  Q108K:K40D	
  holo	
  DS	
  dimer:	
  g.	
  Chains	
  A	
  and	
  F	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  h.	
  Chains	
  D	
  and	
  H.	
  	
  i.	
  Chain	
  L	
  and	
  J.	
  j.	
  Chains	
  B	
  and	
  E	
  k.	
  Chains	
  K	
  and	
  I.	
  l.	
  Chains	
  C	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  G.......................................................................................................................71	
  
	
  
Figure	
  II-­‐29:	
  	
  UV	
  data	
  spectrum	
  for	
  different	
  dimer	
  mutants	
  with	
  all-­‐trans	
  retinal	
  was	
  carried	
  by	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Dr.	
  Santos.	
  Left,	
  UV	
  spectra	
  represents	
  the	
  kinetically	
  formed	
  product	
  between	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hCRBPII	
  dimer	
  and	
  retinal.	
  Right,	
  this	
  panel	
  depicts	
  the	
  formation	
  of	
  both	
  SB	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  PSB	
  as	
  a	
  function	
  of	
  time.	
  ........................................................................................73	
  
	
  
Figure	
  II-­‐31:	
  	
  Circular	
  Dichroism	
  spectra	
  of	
  various	
  hCRBPII	
  mutants	
  as	
  a	
  function	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  temperature.	
  a.	
  Monomer	
  KLY60W.	
  b.	
  Dimer	
  KLY60W.	
  c.	
  Monomer	
  Y60L.	
  d.	
  Dimer	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Y60L.	
  e.	
  Monomer	
  WT	
  hCRBPII.	
  f.	
  Dimer	
  WT	
  hCRBPII.	
  	
  Note,	
  most	
  of	
  the	
  secondary	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structure	
  is	
  preserved	
  in	
  both	
  the	
  monomers,	
  even	
  at	
  the	
  highest	
  temperature	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  achieved.	
  	
  Thus	
  melting	
  curves	
  can	
  only	
  give	
  a	
  lower	
  limit	
  for	
  the	
  TM,	
  consistent	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  the	
  view	
  that	
  the	
  monomers	
  are	
  more	
  stable	
  than	
  the	
  dimers......................79	
  
	
  
Figure	
  III-­‐1:	
  Overlaid	
  of	
  FABP4	
  (2q9s,	
  green)	
  -­‐Linoleic	
  acid	
  (blue),	
  FABP4	
  (2ans,	
  pink)-­‐oleic	
  acid	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  (grey)	
  shows	
  the	
  exit	
  of	
  the	
  oleic	
  acid	
  from	
  the	
  binding	
  pocket...............................96	
  
	
  
Figure	
  III-­‐2:	
  Size	
  Exclusion	
  Chromatogram	
  of	
  human	
  FABP4,	
  which	
  was	
  expressed	
  in	
  E.coli	
  (at	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  30°C	
  as	
  dimer	
  favoring	
  temperature,	
  explained	
  in	
  section	
  III-­‐3,	
  green).	
  SEC	
  profile	
  of	
  	
  
xviii	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  monomer	
  hCRBPII	
  mutant	
  (orange)	
  as	
  our	
  size	
  control............................................98	
  
	
  
Figure	
  III-­‐3:	
  	
  Model	
  for	
  the	
  role	
  of-­‐FABP5	
  and	
  its	
  involvement	
  in	
  regulating..............................99	
  
	
  
Figure	
  III-­‐4	
  	
  Overlaid	
  of	
  DS	
  dimers	
  of	
  holo	
  hFABP5	
  and	
  hCRBPII.	
  HCRBPII	
  (purple,	
  PDB	
  ID	
  4ZH9),	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  BMS-­‐hFABP5	
  (green),	
  AEA-­‐hFABP5	
  (orange).	
  There	
  is	
  a	
  big	
  different	
  orientation	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  two	
  domains	
  upon	
  ligand	
  binding.	
  .........................................................................100	
  
	
  
Figure	
  III-­‐5:	
  a.	
  Overlaid	
  of	
  holo	
  DS	
  dimers	
  of	
  BMS-­‐hFABP5	
  (green),	
  AEA-­‐hFABP5	
  (orange).	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Dramatic	
  change	
  is	
  seen	
  in	
  the	
  relative	
  orientation	
  of	
  two	
  domains	
  in	
  these	
  DS	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  dimers,	
  which	
  is	
  induced	
  by	
  different	
  ligands.	
  b.	
  Overlaid	
  of	
  DS	
  dimers	
  of	
  BMS-­‐	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hFABP5	
  (green),	
  AEA-­‐hFABP5	
  (orange)	
  and	
  holo	
  Q108K:K40D	
  DS	
  dimer	
  of	
  hCRBPII.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  AEA,	
  BMS	
  and	
  retinal	
  are	
  shown	
  as	
  sticks.	
  Between	
  structures	
  of	
  dimer	
  hCRBPII	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  AEA-­‐hFABP5,	
  there	
  is	
  14Å	
  deviation	
  while	
  for	
  BMS-­‐hFABP5	
  dimer	
  there	
  is	
  a	
  larger	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  deviation	
  of	
  17Å.......................................................................................................102	
  
	
  
Figure	
  III-­‐6:	
  a.	
  SEC	
  of	
  WT	
  and	
  Q4A	
  mutant	
  FABP5	
  purification	
  by	
  SEC	
  after	
  E.	
  coli	
  expression.	
  	
  b.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  SDS-­‐PAGE	
  of	
  SEC	
  fractions	
  18-­‐21	
  of	
  WT-­‐hFABP5.	
  c.	
  	
  Native	
  gel	
  of	
  the	
  same	
  SEC	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  fractions.	
  It	
  shows	
  that	
  in	
  the	
  dimer	
  peak,	
  both	
  monomer	
  and	
  dimer	
  species	
  are	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  present.....................................................................................................................104	
  
	
  
Figure	
  III-­‐7:	
  Overlaid	
  of	
  our	
  current	
  monomer	
  structure	
  (chain	
  A	
  in	
  green	
  and	
  chain	
  B	
  in	
  pink)	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  hFABP5	
  monomer	
  (grey,	
  PDB	
  ID	
  4lkp).	
  There	
  is	
  a	
  different	
  conformation	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  between	
  their	
  3rd	
  and	
  4th	
  beta	
  sheet	
  (hinge	
  loop	
  region	
  in	
  DS	
  dimer),	
  which	
  is	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  pointed	
  by	
  arrow.	
  ...................................................................................................105	
  
	
  
Figure	
  III-­‐8:	
  a.	
  Crystal	
  structure	
  of	
  DS	
  dimer	
  of	
  hFABP5	
  of	
  our	
  trial.	
  Chain	
  A	
  in	
  blue,	
  Chain	
  B	
  in	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  cyan.	
  b.	
  Overlaid	
  of	
  DS	
  hFABP5	
  structure	
  	
  (obtained	
  by	
  Nona,	
  cyan)	
  with	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  abovementioned	
  pseudo	
  monomer	
  hFABP5	
  (chain	
  A	
  in	
  pink	
  and	
  chain	
  B	
  in	
  green)c.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Overlay	
  of	
  our	
  determined	
  DS	
  hFABP5	
  structure	
  	
  (blue)	
  with	
  DS	
  dimer	
  of	
  BMS-­‐	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hFABP5	
  (green,	
  4AZM),	
  and	
  AEA-­‐hFABP5	
  (orange,	
  4AZR).	
  	
  This	
  dimer	
  structure	
  is	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  very	
  similar	
  to	
  the	
  AEA-­‐	
  structure	
  and	
  both	
  have	
  different	
  orientation	
  of	
  their	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  domains	
  compared	
  with	
  the	
  BMS-­‐structure............................................................107	
  
	
  
Figure	
  III-­‐9:	
  a.	
  In	
  DS	
  dimer	
  of	
  hFABP5	
  (Chain	
  A	
  in	
  salmon,	
  chain	
  B	
  in	
  cyan,	
  PDB	
  4AZR),	
  there	
  is	
  a	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hydrogen	
  bond	
  between	
  Q65	
  and	
  T57.	
  b.	
  While	
  in	
  monomer	
  structure	
  (purple,	
  PDB	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  ID	
  4LKP)	
  Gln65	
  is	
  pointing	
  out	
  of	
  toward	
  the	
  solvent	
  and	
  T64	
  makes	
  a	
  hydrogen	
  bond	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  to	
  Thr57	
  via	
  water	
  network.	
  Res.	
  57,	
  64	
  and	
  65	
  are	
  represented	
  in	
  sticks.	
  Water	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  molecules	
  are	
  in	
  red.................................................................................................108	
  
	
  
Figure	
  III-­‐10:	
  a.	
  Size	
  exclusion	
  chromatogram	
  of	
  Gln65	
  mutants	
  of	
  hFABP5	
  during	
  purification	
  by	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  SEC.	
  b.	
  Size	
  exclusion	
  chromatogram	
  of	
  Thr64	
  mutants	
  of	
  hFABP5	
  (Thr63E	
  pink,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  T63L	
  green)	
  during	
  purification	
  by	
  SEC..................................................................109	
  
	
  
xix	
  

Figure	
  III-­‐11:	
  Phase	
  relationship	
  in	
  iLBP	
  family...........................................................................111	
  
	
  
Figure	
  IV-­‐1:	
  a.	
  Polar	
  residues	
  around	
  the	
  ligand	
  mutated	
  in	
  order	
  to	
  remove	
  polarity	
  from	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  binding	
  pocket.	
  Coordinates	
  obtained	
  from	
  PDB	
  4EXZ	
  (hCRBPII-­‐Q108K:K40L/retinal).	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  b.	
  Both	
  retinal	
  and	
  ThioFluor	
  bind	
  and	
  lay	
  in	
  the	
  same	
  region	
  of	
  the	
  binding	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  pocket.	
  (the	
  Thioflour	
  coordinates	
  are	
  from	
  	
  Q108K:K40L:	
  T51V:T53S:	
  R58Y	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  structure,	
  will	
  be	
  discussed	
  later	
  in	
  the	
  IV-­‐3	
  section.	
  c.	
  The	
  chemical	
  Structure	
  of	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  ThioFluor	
  ,	
  which	
  	
  was	
  synthesized	
  by	
  Dr.	
  Santos....................................................129	
  
	
  
Figure	
  IV-­‐2:	
  	
  a.	
  Water	
  network	
  hydrogen	
  bonding	
  between	
  T51	
  and	
  T53.	
  b.	
  Water	
  mediated	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hydrogen	
  bonding	
  between	
  Q38	
  and	
  Q128.	
  (PDB	
  4EXZ,	
  hCRBPII-­‐	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Q108K:K40L/retinal,	
  Retinal	
  is	
  represented	
  in	
  green	
  sticks)...................................130	
  
	
  
Figure	
  IV-­‐3:	
  a.	
  Crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  and	
  space-­‐filling	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  representation	
  obtained	
  by	
  Dr.	
  Nosrati.	
  R58W	
  Is	
  packed	
  with	
  the	
  N,N-­‐dimethyl	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  amino	
  moiety	
  of	
  the	
  ThioFlour	
  tail.	
  b.	
  The	
  water	
  mediated	
  hydrogen	
  bonding	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  between	
  T53S	
  and	
  R58W	
  is	
  highlighted..................................................................133	
  
	
  
Figure	
  IV-­‐4:	
  	
  a.	
  Overlay	
  of	
  Crystal	
  structures	
  of	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  (cyan)	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  Q108K:K40L:T51V:T53S:R58Y/ThioFluor	
  (green).	
  b.	
  It	
  shows	
  Water	
  mediated	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  hydrogen	
  bonding	
  between	
  R58Y	
  and	
  	
  T29............................................................134	
  
	
  
Figure	
  IV-­‐5:	
  	
  Overlay	
  of	
  Crystal	
  structures	
  of	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  (cyan)	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  with	
  Q108K:K40L:T51V:T53S:R58Y/ThioFluor	
  (green)	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Q108K:K40L:T53A:R58F/ThioFluor	
  	
  (salmon)........................................................135	
  
	
  
Figure	
  IV-­‐6:	
  The	
  water	
  mediated	
  hydrogen	
  bonding	
  between	
  Tyr19	
  to	
  the	
  thiophene	
  sulfur	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  ThioFluor	
  highlighted	
  in	
  the	
  Crystal	
  structure	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  obtained	
  by	
  Dr.	
  Nosrati.	
  ......................	
  136	
  
	
  
Figure	
  IV-­‐7:	
  a.	
  Crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor.	
  b.	
  Water	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  network	
  is	
  abolished	
  in	
  this	
  structure.	
  	
  ThioFluor	
  and	
  residues	
  around	
  it	
  are	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  represented	
  sticks....................................................................................................137	
  
	
  
Figure	
  IV-­‐8:	
  Crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  obtained	
  by	
  Alireza.	
  	
  ThioFluor	
  in	
  cyan	
  and	
  residues	
  around	
  it	
  in	
  green	
  are	
  	
  shown	
  in	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  sticks........................................................................................................................138	
  
	
  
Figure	
  IV-­‐9:	
  Mutated	
  residues	
  which	
  discussed	
  in	
  this	
  chapter	
  (show	
  in	
  blue)	
  in	
  an	
  attempt	
  to	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  encapsulate	
  the	
  binding	
  cavity.	
  Crystal	
  structure	
  is	
  of	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor.....................................................139	
  
	
  
Figure	
  IV-­‐10:	
  Crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor.	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  ThioFluor	
  and	
  highlighted	
  residues	
  are	
  represented	
  as	
  sticks...............................140	
  
xx	
  

Figure	
  V-­‐1:	
  Schematic	
  view	
  of	
  the	
  BE	
  isozymes	
  and	
  SS	
  isozymes	
  in	
  the	
  amylopectin	
  cluster....159	
  
Figure	
  V-­‐2:	
  Overall	
  structure	
  of	
  rice	
  Branching	
  Enzyme	
  I	
  in	
  complex	
  with	
  maltododecaose	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  (M12).	
  The	
  N-­‐terminal	
  domain	
  is	
  shown	
  in	
  green,	
  CBM48,	
  pink,	
  center	
  catalytic	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  domain,	
  cyan,	
  and	
  the	
  C-­‐terminal	
  α-­‐amylase	
  C	
  domain	
  in	
  blue.	
  Carbohydrates	
  are	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  represented	
  in	
  sticks	
  and	
  colored	
  by	
  atom	
  type,	
  carbons	
  in	
  yellow	
  and	
  oxygens	
  in	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  red.	
  One	
  oligosaccharide,	
  M12,	
  binds	
  exclusively	
  into	
  the	
  catalytic	
  domain	
  and	
  hangs	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  over	
  the	
  catalytic	
  groove	
  without	
  reaching	
  inside	
  (site	
  II),	
  while	
  the	
  five	
  glucose	
  units	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  visible	
  for	
  the	
  second	
  molecules	
  (site	
  I),	
  between	
  three	
  domains:	
  the	
  N-­‐terminal,	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  carbohydrate	
  binding	
  module,	
  and	
  the	
  catalytic	
  domain.	
  .......................................161	
  
	
  
Figure	
  V-­‐3:	
  Surface	
  depiction	
  of	
  RBEI	
  in	
  complex	
  with	
  M12.	
  At	
  the	
  center	
  of	
  this	
  groove,	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  residues	
  involved	
  in	
  catalysis	
  are	
  shown	
  in	
  blue.	
  These	
  catalytic	
  residues,	
  Y235,	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  D270,	
  H275,	
  R342,	
  D344,	
  E399,	
  H467	
  and	
  D468	
  according	
  to	
  RBEI	
  sequence	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  numbering,	
  (2)	
  were	
  predicted	
  based	
  on	
  biochemical	
  and	
  structural	
  data	
  of	
  α-­‐	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  amylase.	
  (3,	
  4)	
  ............................................................................................................162	
  
	
  
Figure	
  V-­‐4:	
  Binding	
  site	
  I:	
  detailed	
  interactions	
  between	
  the	
  oligosaccharide	
  and	
  RBEI.	
  The	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  protein	
  atoms	
  are	
  colored	
  by	
  type:	
  C	
  in	
  blue	
  marine,	
  O	
  in	
  red	
  and	
  N	
  in	
  dark	
  blue.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  M12	
  atoms	
  are	
  also	
  colored	
  by	
  type	
  with	
  C	
  in	
  yellow	
  and	
  O	
  in	
  red.	
  Glucose	
  units	
  are	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  numbered	
  in	
  red.	
  Hydrogen	
  bonds	
  are	
  shown	
  in	
  dotted	
  black	
  lines.	
  Water	
  molecules	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  interacting	
  in	
  this	
  site	
  are	
  represented	
  in	
  spheres	
  and	
  colored	
  in	
  cyan...................163	
  
	
  
Figure	
  V-­‐5:	
  	
  Binding	
  site	
  II,	
  detailed	
  interactions	
  between	
  the	
  oligosaccharide	
  and	
  RBEI.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Numbers	
  for	
  glucose	
  units,	
  and	
  the	
  protein	
  atoms	
  are	
  colored.	
  Hydrogen	
  bonds	
  and	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  water	
  molecules	
  are	
  also	
  represented.	
  Hydrogen	
  bonds	
  are	
  shown	
  in	
  dotted	
  black	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  lines.	
  .......................................................................................................................164	
  
	
  
Figure	
  V-­‐6:	
  Overall	
  structure	
  of	
  RBEI.	
  Active	
  site	
  residues	
  are	
  shown	
  in	
  blue	
  sticks.	
  Two	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  oligosaccharide	
  molecules	
  are	
  in	
  yellow	
  sticks.	
  Oxygen	
  atoms	
  are	
  in	
  red.	
  So	
  far,	
  we	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  have	
  not	
  identified	
  any	
  glucose	
  units	
  close	
  and	
  around	
  the	
  active	
  site	
  yet.............166	
  
	
  
Figure	
  V-­‐7:	
  Absorbance	
  versus	
  Time	
  for	
  the	
  D344A	
  mutation	
  of	
  truncated	
  RBEI.The	
  graph	
  is	
  the	
  	
  
average	
  for	
  three	
  trials.	
  	
  Slope/Protein	
  indicates	
  the	
  activity	
  (U/mg)	
  of	
  the	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
enzyme......................................................................................................................167	
  
	
  
Figure	
  V-­‐8:	
  Detailed	
  interactions	
  between	
  the	
  Glu7	
  in	
  binding	
  site	
  II	
  and	
  residue	
  E534	
  of	
  RBEI.	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  All	
  interactions	
  between	
  them	
  are	
  in	
  between	
  2-­‐3.5	
  Å.	
  	
  Numbers	
  for	
  glucose	
  units,	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  and	
  the	
  protein	
  atoms	
  are	
  colored.	
  Hydrogen	
  bonds	
  and	
  water	
  molecules	
  are	
  also	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  represented.	
  Hydrogen	
  bonds	
  are	
  shown	
  in	
  dotted	
  black	
  lines.	
  .............................168	
  
	
  
Figure	
  V-­‐9:	
  	
  Absorbance	
  versus	
  Time	
  for	
  the	
  E534A	
  mutation	
  of	
  truncated	
  RBEI:	
  The	
  graph	
  is	
  the	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  average	
  for	
  three	
  trials.	
  The	
  decrease	
  in	
  absorbance	
  is	
  linear	
  for	
  the	
  first	
  5	
  min	
  then	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  reaches	
  a	
  plateau	
  due	
  to	
  the	
  high	
  activity	
  of	
  the	
  enzyme......................................169	
  
xxi	
  

KEY	
  TO	
  SYMBOLS	
  AND	
  ABBREVIATIONS	
  

Å	
  

	
  

Angstrom	
  

PDB	
  ID	
  	
  

Protein	
  Data	
  Bank	
  Identifier	
  

PSB	
  

	
  

Protonated	
  Schiff	
  Base	
  

SB	
  

	
  

Schiff	
  Base	
  

iLBP	
  	
   	
  

intracellular	
  Lipid	
  Binding	
  Proteins	
  

hCRBPII	
  	
  

human	
  Cellular	
  Retinol	
  Binding	
  ProteinII	
  

hCRABPII	
  	
  

human	
  Cellular	
  Retinoic	
  Acid	
  Binding	
  ProteinII	
  

hFABP4	
  

human	
  Fatty	
  Acid	
  Binding	
  Protein4	
  

hFABP5	
  

human	
  Fatty	
  Acid	
  Binding	
  Protein5	
  

σ	
  

	
  

Sigma	
  

UV	
  

	
  

Ultra	
  Violet	
  

WT	
  	
  

	
  

Wild-­‐Type	
  

IPTG	
   	
  

Isopropyl	
  &-­‐D-­‐1-­‐thiogalactopyranoside	
  

PCR	
  

	
  

Polymerase	
  Chain	
  Reaction	
  

PEG	
  

	
  

Polyethylene	
  glycol	
  

Vis	
  

	
  

Visible	
  

Kd	
  

	
  

Dissociation	
  constant	
  

SDS-­‐PAGE	
  

Sodium	
  dodecyl	
  sulfate	
  polyacrylamide	
  gel	
  electrophoresis	
  

KDa	
  

Kilo	
  Dalton	
  

	
  

	
  
xxii	
  

E.	
  Coli's	
  

Escherichia	
  coli	
  

min	
  

	
  

minute	
  

S	
  

	
  

Second	
  

Îľ	
  

	
  

Extinction	
  coefficient	
  

QY	
  

	
  

Quantum	
  Yield	
  

Îťmax	
  

	
  

Maximum	
  wavelength	
  

Îťex	
  

	
  

Excitation	
  wavelength	
  

Îťem	
  

	
  

Emission	
  Wavelength	
  

mM	
  

	
  

Milimolar	
  	
  

ÎźM	
  	
  

	
  

Micromolar	
  	
  

nM	
  	
  

	
  

Nanomolar	
  	
  

mol	
  	
   	
  

mole	
  

mmol	
  	
   	
  

millimole	
  

mg	
  	
  

	
  

milligram	
  

ml	
  	
  

	
  

milliliter	
  

DNA	
   	
  

deoxyribonucleic	
  acid	
  

dNTP	
  	
   	
  

deoxynucleotide	
  triphosphates	
  	
  

rpm	
  	
   	
  

rotation	
  per	
  minute	
  

°C	
  	
  

	
  

degrees	
  of	
  centigrade	
  

K	
  	
  

	
  

degrees	
  of	
  kelvin	
  

pH	
  	
  

	
  

Logarithmic	
  scale	
  of	
  hydrogen	
  ion	
  activity	
  	
  

NaOH	
  	
   	
  

Sodium	
  hydroxide	
  
xxiii	
  

Equiv	
  	
   	
  

equivalent	
  

RMSD	
  	
   	
  

root	
  mean	
  square	
  deviation	
  

Ala,	
  A	
  	
   	
  

Alanine	
  

Arg,	
  R	
  	
   	
  

Argannine	
  

Asn,	
  N	
   	
  

Aspargine	
  

Asp,	
  D	
  	
  	
  

Aspartate	
  

Cys,	
  C	
  	
   	
  

Cysteine	
  

Gln,	
  Q	
  	
  	
  

Glutamine	
  

Glu,	
  E	
  	
   	
  

Glutamate	
  

His,	
  H	
  	
   	
  

Histitidine	
  

Ile,	
  I	
  	
   	
  

Isoleucine	
  

Leu,	
  L	
  	
   	
  

Leucine	
  

Lys,	
  K	
  	
   	
  

Lysine	
  

Met,	
  M	
  	
  

Methionine	
  

Phe,	
  F	
  	
   	
  

Phenylalanine	
  

Pro,	
  P	
  	
   	
  

Proline	
  

Ser,	
  S	
  	
   	
  

Serine	
  

Thr,	
  T	
  	
   	
  

Threonine	
  

Trp,	
  W	
  	
  

Tryptophan	
  

Tyr,	
  Y	
  	
   	
  

Tyrosine	
  

Val,	
  V	
   	
  

Valine

	
  

xxiv	
  

DS	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  

Domain	
  Swapping	
  	
  

LB	
  	
  

Luria	
  Broth	
  	
  

	
  

R-­‐factor	
  	
  

reliability	
  factor	
  	
  

RT	
  	
  

Room	
  temperature	
  

	
  

RMSD	
  	
   	
  

root-­‐mean-­‐square	
  deviation	
  	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

	
  

xxv	
  

Chapter	
  I:	
  Structural	
  Studies	
  of	
  Human	
  Cellular	
  Retinol	
  Binding	
  Protein	
  
II	
  (hCRBPII)	
  bound	
  to	
  Retinol	
  and	
  Retinal.	
  
	
  	
  	
  	
  	
  

I-­‐1	
  Introduction	
  
	
  

Vitamin	
   A	
   (retinol)	
   is	
   an	
   essential	
   micronutrient	
   that	
   plays	
   a	
   key	
   role	
   in	
   vision,	
   cell	
  

growth,	
   differentiation	
   metabolism	
   etc.	
   (5)	
   Because	
   of	
   its	
   low	
   solubility	
   in	
   aqueous	
   medium,	
  
retinol	
  is	
  bound	
  by	
  specific	
  binding	
  proteins	
  in	
  body	
  fluids	
  and	
  within	
  the	
  cell	
  for	
  its	
  transport,	
  
bioavailability	
   and	
   stability.	
   Retinol	
   is	
   converted	
   to	
   retinal,	
   which	
   is	
   the	
   essential	
   molecule	
   of	
  
vision,	
  and	
  also	
  to	
  retinoic	
  acid,	
  which	
  is	
  essential	
  in	
  cell	
  differentiation.	
  (6)	
  
In	
  different	
  tissues,	
  different	
  binding	
  proteins	
  play	
  important	
  roles	
  in	
  transport,	
  storage	
  
and	
   metabolism	
   of	
   vitamin	
   A,	
   such	
   as	
   Retinol	
   Binding	
   Proteins	
   (RBP),	
   Cellular	
   Retinoic	
   Acid	
  
Binding	
   Proteins	
   (CRABP),	
   Cellular	
   Retinol	
   Binding	
   Protein	
   (CRBP),	
   Cellular	
   Retinaldehyde	
  
	
  
Binding	
   Protein	
   (CRALBP)	
   and	
   Interphotoreceptor	
   Retinol	
   Binding	
   Protein	
   (IRBP). In	
   all	
   those	
  
proteins,	
  the	
  hydrophobic	
  ligand	
  and	
  protein	
  have	
  non-­‐covalent	
  interactions.(7,	
  8)

	
  

CRBPs	
   belong	
   to	
   the	
   family	
   of	
   intracellular	
   lipid	
   binding	
   proteins	
   (iLBP’s).	
   	
   iLBPs	
   are	
   a	
  
sub-­‐family	
  of	
  the	
  large	
  family	
  of	
  calycins	
  that,	
  in	
  addition	
  to	
  the	
  iLBP’s,	
  include	
  the	
  avidins	
  and	
  
lipocallins.	
  	
  All	
  the	
  Calycins	
  have	
  a	
  beta-­‐sandwich	
  structure	
  in	
  common	
  and	
  all	
  are	
  chaperones	
  
for	
   their	
   hydrophobic	
   ligands,	
   which	
   bind	
   in	
   their	
   relatively	
   large	
   binding	
   cavities,	
   though	
  
sequence	
  homology	
  between	
  the	
  subfamilies	
  is	
  less	
  than	
  10%.	
   (6,	
  9-­‐13)	
  ILBPs	
  are	
  relatively	
  small	
  
proteins	
   (126-­‐140	
   amino	
   acid)	
   found	
   in	
   the	
   cytosols	
   of	
   fish,	
   amphibians,	
   reptiles,	
   birds	
   and	
  
mammals.	
   They	
   responsible	
   for	
   shuttling	
   insoluble	
   hydrophobic	
   molecules,	
   including	
   retinal,	
  
retinoic	
   acid,	
   various	
   long	
   chain	
   fatty	
   acids,	
   eicosinoids,	
   cholesterol	
   and	
   hemes,	
   throughout	
   the	
  

1	
  

cell.(6,	
   10-­‐13)	
   Their	
   structure	
   consists	
   of	
   a	
   ten-­‐stranded	
  𝛽 -­‐barrel,	
   and	
   two	
   short	
  𝛼-­‐helices	
   that	
  
cover	
  the	
  binding	
  pocket,	
  with	
  the	
  ligand	
  buried	
  deeply	
  within	
  the	
  binding	
  site.	
  Basically,	
  two	
  
short	
  ι	
  helixes	
  cover	
  the	
  binding	
  pocket	
  like	
  a	
  lid,	
  isolating	
  the	
  ligand	
  from	
  bulk	
  solution	
  and	
  the	
  
hydrophobic	
  ligand	
  is	
  deeply	
  buried	
  inside	
  the	
  binding	
  pocket.	
   (14)(Figure	
   I-­‐1a)	
   This	
  family	
  also	
  
includes	
  the	
  Cellular	
  Retinoic	
  Acid	
  Binding	
  Proteins	
  (CRABP),	
  and	
  the	
  Fatty	
  Acid	
  Binding	
  Proteins	
  
(FABP)	
  with	
  quite	
  similar	
  structures.(8)	
  
	
  

b.

a.	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   I-­‐1:	
   a.	
   An	
   overlay	
   of	
   the	
   structures	
   of	
   hCRBPII	
   (green	
   and	
   cyan),	
  
hCRBPIII	
  (1GGL,	
  salmon),	
  hCRBPIV	
  (1LPJ,	
  yellow)	
  and	
  hCRBPI	
  (1KGL,	
  pink)	
  
b.	
   An	
   overlay	
   of	
   the	
   structures	
   of	
   apo	
   hCRBPII	
   (2RCQ,	
   gray),	
   wt	
   hCRBPII	
  
bound	
   to	
   all-­‐trans-­‐retinal	
   (4QYP,	
   green)	
   and	
   to	
   all-­‐trans-­‐retinol	
   (4QYN,	
  
cyan).	
  
There	
  are	
  four	
  identified	
  isoforms	
  for	
  CRBP	
  (CRBPI,	
  II,	
  III	
  and	
  IV),	
  among	
  them	
  only	
  CRBPI	
  

and	
  II	
  have	
  been	
  shown	
  to	
  be	
  involved	
  in	
  the	
  metabolism	
  and	
  transport	
  of	
  intracellular	
  retinol	
  
so	
  far.	
  HCRBPI	
  is	
  found	
  predominantly	
  in	
  the	
  liver	
  and	
  kidney	
  while	
  hCRBPII	
  is	
  more	
  prevalent	
  in	
  
the	
   small	
   intestine.	
   (5,	
   15,	
   16)It	
   was	
   reported	
   that	
   hCRBPII	
   regulates	
   retinoid	
   metabolism	
   in	
  
intestine	
  and	
  facilitates	
  the	
  reduction	
  of	
  retinal	
  to	
  retinol	
  and	
  the	
  subsequent	
  esterification	
  of	
  
retinal	
   to	
   retinyl	
   ester.	
   HCRBPIII	
   mRNA	
   is	
   expressed	
   in	
   kidney	
   and	
   liver	
   thus	
   suggesting	
   an	
  
impotent	
   role	
   as	
   an	
   intracellular	
   mediator	
   of	
   retinol	
   metabolism	
   in	
   these	
   tissues.	
   	
   HCRBPIV	
  
2	
  

mRNA	
   is	
   expressed	
   primarily	
   in	
   heart,	
   and	
   transverse	
   colon.	
   However,	
   physiological	
   roles	
   of	
  
newly	
  identified	
  hCRBPIII	
  and	
  IV	
  are	
  still	
  unknown.(8,	
  17)	
  

	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   I-­‐2:	
   The	
   multiple	
   sequence	
   alignment	
   between	
   human,	
   rat	
   CRBPs	
  
and	
  zebra	
  fish	
  CRBP.	
  

	
  

CRBPI	
  and	
  II	
  share	
  56%	
  sequence	
  identity	
  and	
  both	
  proteins	
  can	
  bind	
  all-­‐trans-­‐retinol,	
  all	
  
trans-­‐retinal	
  and	
  13-­‐cis-­‐retinol,	
  while	
  neither	
  bind	
  9-­‐cis-­‐retinol	
  though	
  CRBPII	
  is	
  found	
  to	
  be	
  the	
  
more	
  selective	
  for	
  all-­‐trans-­‐retinoids.	
  (18-­‐20)	
  (Figure	
  I-­‐2)	
  As	
  the	
  first	
  crystal	
  structures	
  of	
  CRBP,	
  we	
  
can	
  point	
  to	
  the	
  apo	
  and	
  all-­‐trans-­‐retinol	
  bound	
  structure	
  of	
  rat	
  CRBPII,	
  which	
  was	
  published	
  in	
  
1993.(21-­‐23)	
   Comparison	
   of	
   the	
   two	
   structures	
   showed	
   that	
   no	
   significant	
   differences	
   were	
  
observed	
  upon	
  ligand	
  binding.	
  Later	
  on,	
  the	
  crystal	
  structures	
  of	
  human	
  CRBPII	
  and	
  a	
  zebra	
  fish	
  
CRBP,	
   both	
   bound	
   with	
   all-­‐trans-­‐retinol,	
   were	
   also	
   reported.	
   (21,	
  22)However,	
   in	
   both	
   cases	
   the	
  
electron	
   density	
   of	
   the	
   end	
   of	
   the	
   chromophore	
   was	
   not	
   clear.	
   The	
   ligand	
   was	
   modeled	
   as	
   a	
  
mixture	
  of	
  retinol,	
  and	
  a	
  retinol	
  derivative	
  whereas	
  the	
  C13-­‐C14	
  bond	
  is	
  a	
  single	
  bond.	
  (2RCT,	
  
1KQW)	
  
In	
   our	
   lab	
   and	
   in	
   collaboration	
   with	
   prof.	
   Borhan’s	
   group,	
   we	
   were	
   the	
   first	
   group	
   to	
  
achieve	
   the	
   structure	
   of	
   an	
   all-­‐trans-­‐retinal	
   bound	
   CRBP	
   (hCRBPII)	
   by	
   Dr.	
   Nossoni.	
   	
   We	
   also	
  
3	
  

were	
   the	
   first	
   to	
   show	
   that	
   a	
   retinoid	
   derivative	
   virtually	
   identical	
   to	
   that	
   seen	
   in	
   previously	
  
published	
   CRBPII	
   structures	
   can	
   be	
   generated	
   by	
   X-­‐ray	
   radiation	
   damage,	
   and	
   this	
   process	
   is	
  
dependent	
  on	
  the	
  wavelength	
  of	
  the	
  X-­‐ray	
  radiation.	
  In	
  addition,	
  we	
  finally	
  determined	
  the	
  first	
  
bonafide	
  structure	
  of	
  retinol-­‐bound	
  hCRBPII.(6)	
  	
  (Figure	
  I-­‐1b)	
  

I-­‐2	
  The	
  affinity	
  of	
  hCRBPII	
  for	
  retinol	
  and	
  retinal	
  
	
  The	
   affinity	
   of	
   rat	
   CRBPII	
   for	
   both	
   retinol	
   and	
   retinal	
   has	
   been	
   measured	
   using	
   both	
   a	
  
fluorescence	
   quenching	
   assay	
   and	
   Fluorine	
   NMR.	
   The	
   most	
   recent	
   study	
   showed	
   rat	
   CRBPII	
  
binds	
   to	
   retinol	
   and	
   retinal	
   with	
   similar	
   affinity	
   (a	
   Kd	
   of	
   about	
   10	
   nM).(20,	
  24,	
  25)	
   As	
   far	
   as	
   we	
   knew,	
  
no	
   affinity	
   measurements	
   for	
   hCRBPII	
   had	
   been	
   reported,	
   and	
   then,	
   we	
   decided	
   to	
   measure	
   its	
  
affinity	
  to	
  both	
  retinal	
  and	
  retinol	
  using	
  a	
  Tryptophan	
  fluorescence-­‐quenching	
  assay	
  	
  (Described	
  
in	
  Experimental	
  Section).	
  	
  While	
  the	
  binding	
  of	
  retinal	
  was	
  significantly	
  weaker	
  compared	
  with	
  
that	
   most	
   recently	
   reported	
   for	
   the	
   rat	
   protein,	
   130¹10	
   nM	
   VS	
   10nM	
   in	
   rat	
   CRBPII,	
   Figure	
   I-­‐
3d)(24),	
  hCRBPII	
  showed	
  significantly	
  higher	
  affinity	
  for	
  retinol,	
  with	
  a	
  lower	
  limit	
  of	
  2.2¹7.9	
  nM	
  
(Figure	
   I-­‐3a).	
   The	
   binding	
   constant	
   appeared	
   to	
   be	
   too	
   high	
   for	
   accurate	
   measurement	
   using	
  
our	
  fluorescence	
  assay.	
  	
  This	
  seems	
  to	
  indicate	
  a	
  potential	
  difference	
  in	
  the	
  interaction	
  of	
  rat	
  
versus	
  human	
  CRBP	
  for	
  these	
  ligands.	
  The	
  ligand	
  is	
  stabilized	
  within	
  binding	
  pocket,	
  allowing	
  it	
  
to	
  preserve	
  its	
  yellow	
  colour	
  at	
  room	
  temperature	
  and	
  in	
  light	
  for	
  more	
  than	
  three	
  days,	
  while	
  
in	
  ethanol	
  solution	
  a	
  quarter	
  of	
  the	
  chromophore	
  is	
  degraded	
  after	
  4	
  hours	
  when	
  exposed	
  to	
  
air.	
  

4	
  

	
  

a	
  

Binding	
  A ffinity	
  of	
  
WT	
  hCRBPII	
  to	
  A ll-­‐
Trans-­‐Retinol	
  

b	
  

	
  
Kd=	
  2.2nM¹7.9	
  

	
  

Binding	
  A ffinity	
  of	
  
T51I-­‐hCRBPII	
  to	
  All-­‐
Trans-­‐Retinol	
  
Kd=	
  2.2nM¹6.7	
  

	
  

	
  
	
  

c	
  
	
  

Binding	
  A ffinity	
  of	
  
T51I-­‐hCRBPII	
  to	
  All-­‐
Trans-­‐Retinal	
  
Kd=46nM Âą	
  10	
  

	
  

	
  

d	
  

Binding	
  A ffinity	
  of	
  WT	
  
hCRBPII	
  to	
  All-­‐Trans-­‐
Retinal	
  
Kd=	
  130nM ¹	
  14	
  

	
  

	
  
	
  
Figure	
   I-­‐3:	
   Ligand	
   binding	
   measurements	
   using	
   the	
   fluorescence	
   quenching	
   of	
  
tryptophan	
   (excitation	
   at	
   280	
   nm,	
   emission	
   at	
   350	
   nm).	
   a.	
   Titration	
   of	
   WT-­‐hCRBPII	
  
with	
   all-­‐trans-­‐retinol.	
   b.	
   Titration	
   of	
   hCRBPII	
   T51I	
   mutant	
   with	
   all-­‐trans-­‐retinol.	
   c.	
  
Titration	
   of	
   hCRBPII	
   T51I	
   mutant	
   with	
   all-­‐trans-­‐retinal.	
   	
   d.	
   Titration	
   of	
   WT-­‐hCRBPII	
  
with	
  all-­‐trans-­‐retinal.	
  	
  
	
  
	
  

I-­‐3	
  Structure	
  of	
  the	
  retinal-­‐bound	
  hCRBPII	
  complex	
  
Retinal-­‐bound	
   hCRBPII	
   crystallizes	
   in	
   the	
   P1	
   space	
   group	
   with	
   four	
   molecules	
   per	
  
asymmetric	
   unit	
   (Described	
   in	
   experimental	
   section,	
   Table	
   I-­‐1).(6)	
   Electron	
   density	
  
corresponding	
  to	
  the	
  entire	
  ligand	
  was	
  very	
  clear	
  in	
  only	
  one	
  of	
  the	
  four	
  chains	
  (chain	
  B).	
  	
  In	
  
chain	
   C,	
   density	
   for	
   retinal	
   is	
   well	
   defined	
   for	
   all	
   atoms	
   but	
   C15	
   and	
   the	
   hydroxyl	
   group.	
   	
   In	
  
molecule	
  A,	
  electron	
  density	
  is	
  relatively	
  well-­‐defined	
  for	
  the	
  carbonyl	
  group	
  and	
  C11-­‐C15,	
  but	
  
is	
   weaker	
   for	
   the	
   rest	
   of	
   the	
   molecule.	
   Scattered,	
   uninterruptable	
   electron	
   density	
   is	
   seen	
   in	
  
5	
  

molecule	
   D.	
   The	
   overall	
   structure	
   of	
   the	
   protein	
   is	
   little	
   changed	
   either	
   in	
   the	
   apo	
   state,	
   or	
  
when	
  bound	
  to	
  either	
  retinal	
  or	
  retinol.	
  (Figure	
  I-­‐1a,	
  1b)	
  The	
  retinal	
  is	
  essentially	
  identical	
  in	
  all	
  
three	
  molecules	
  where	
  the	
  electron	
  density	
  is	
  clear	
  (Figure	
  I-­‐4a).	
  	
  
The	
  β-­‐ionone	
  ring	
  adopts	
  the	
  6-­‐s-­‐trans	
  conformation	
  in	
  both	
  ligands.	
  Two	
  hydrophobic	
  
residues,	
   Phe16	
   and	
   Leu77,	
   surround	
   the	
   ionone	
   ring	
   of	
   the	
   ligand,	
   locking	
   the	
   chromophore	
   in	
  
position	
   (Figure	
  I-­‐4b,	
  4c).	
   The	
   distances	
   between	
   the	
   C5	
   methyl	
   group	
   of	
   retinal	
   and	
   Leu77	
   and	
  
Phe16	
  are	
  about	
  3.7Å	
  and	
  3.8Å	
  respectively.	
  (Figure	
  I-­‐4b)	
  	
  	
  
Also,	
  the	
  carbonyl	
  group	
  of	
  retinal	
  makes	
  a	
  water	
  mediated	
  interaction	
  with	
  Gln108	
  and	
  
Lys40.	
  The	
  side	
  chain	
  of	
  Gln108	
  is	
  fixed	
  in	
  position	
  by	
  a	
  small	
  water	
  network,	
  including	
  the	
  main	
  
chain	
  carbonyls	
  of	
  Thr1	
  and	
  Asp91,	
  and	
  the	
  side	
  chains	
  of	
  Gln4,	
  and	
  Gln108	
  (Figure	
  I-­‐4a).	
  This	
  
network	
   is	
   unique	
   to	
   hCRBPII	
   and	
   is	
   not	
   seen	
   in	
   zebra	
   fish	
   CRBP,	
   rat	
   CRBPII	
   and	
   hCRBPI	
   (PDB	
  
entry	
  1KGL),	
  hCRBPIII	
  (PDB	
  entry	
  1GGL)	
  and	
  hCRBPIV	
  (PDB	
  entry	
  1LPJ).	
  (14)	
  
	
  
	
  

	
  

	
  
	
  
	
  

	
  

	
  
	
  
	
  

6	
  
	
  

	
  
	
  

a.	
  
C14

	
  
	
  
	
  
	
  

C15

C13

C12
C11

b.	
  

	
  
	
  

	
  
	
  
	
  
	
  

c.	
  

	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   I-­‐4:	
  Retinal	
  binding	
  in	
  hCRBPII.	
  a.	
  An	
  overlay	
  of	
   retinal	
  from	
  mol	
  A	
  (cyan),	
  mol	
  B	
   (blue)	
  
and	
   Mol	
  C	
  (pink),	
  showing	
  the	
  similarity	
  in	
  binding	
  in	
  the	
  three.	
  	
   b.	
  Stick	
  representation	
  of	
  
the	
  interaction	
  between	
  	
  Phe16,	
  Leu77	
  (both	
  with	
  green	
  carbons,	
  N	
  blue	
  and	
  O,	
  red)	
   and	
  the	
  
retinal	
  ionone	
  ring	
  (blue	
  carbons,	
  Mol	
  B).	
  c.	
  Space	
  filling	
  depiction	
  of	
  b.	
  	
  

7	
  

In	
   this	
   case,	
   water	
   must	
   act	
   as	
   hydrogen	
   bond	
   donor	
   to	
   the	
   two	
   main-­‐chain	
   carbonyl	
   groups	
  
(Thr1	
  and	
  Asp91),	
  only	
  the	
  lone	
  pairs	
  on	
  the	
  oxygen	
  of	
  water	
  are	
  left	
  to	
  have	
  hydrogen	
  bonding	
  
to	
   the	
   amide	
   nitrogen	
   of	
   Gln4,	
   defining	
   its	
   orientation.	
   Therefore	
   it	
   is	
   the	
   carbonyl	
   oxygen	
   of	
  
Gln4	
   that	
   makes	
   a	
   hydrogen	
   bond,	
   necessarily	
   with	
   the	
   amide	
   of	
   Gln108,	
   defining	
   its	
  
orientation	
   as	
   well.	
   The	
   orientation	
   of	
   Gln108,	
   with	
   the	
   carbonyl	
   group	
   pointing	
   toward	
   the	
  
carbonyl	
   of	
   the	
   retinal,	
   abolishing	
   the	
   direct	
   hydrogen	
   bond	
   to	
   the	
   retinal	
   carbonyl	
   oxygen	
   and	
  
necessitates	
  the	
  water-­‐mediated	
  hydrogen	
  bond.	
  The	
  hydrogen	
  bond	
  between	
  this	
  water	
  and	
  
the	
  retinal	
  carbonyl	
  is	
  2.4Å,	
  which	
  is	
  an	
  ideal	
  distance	
  for	
  a	
  strong	
  low-­‐barrier	
  hydrogen	
  bond	
  
(Figure	
   I-­‐5	
   and	
   Figure	
   I-­‐9c).	
  	
  This	
  water	
  makes	
  an	
  additional	
  hydrogen	
  bond	
  with	
  the	
  ε-­‐amino	
  
group	
  of	
  Lys40,	
  and	
  is	
  fixed	
  in	
  space	
  by	
  two	
  hydrogen	
  bonds	
  in	
  the	
  binding	
  pocket.	
  	
  
	
  

Retinal
Gln4

3.0Å

	
  

2.4Å

Thr1

2.5Å

2.5Å

	
  

2.8Å

2.9Å

Gln108

2.7Å

Asp91

Lys40

	
  
	
  

Figure	
  I-­‐5:	
  The	
  water	
  mediated	
  interaction	
  of	
  the	
  retinal	
  hydroxyl	
  
group	
   in	
   hCRBPII.	
   	
   Atoms	
   colored	
   by	
   type,	
   hydrogen	
   bonded	
  
distances	
  are	
  shown.	
  	
  	
  

I-­‐4	
  Wavelength-­‐dependent	
  damage	
  of	
  the	
  retinol	
  in	
  hCRBPII	
  
As	
  previously	
  mentioned,	
  the	
  structures	
  of	
  the	
  CRBPII-­‐retinol	
  complex	
  from	
  rat,	
  human	
  
and	
  zebra	
  fish	
  have	
  already	
  been	
  determined.	
  But	
  the	
  presence	
  of	
  the	
  retinoid	
  derivative	
  that	
  

8	
  

also	
   occupies	
   the	
   binding	
   site	
   in	
   human	
   and	
   zebra	
   fish	
   structures	
   made	
   our	
   interpretation	
  
complicated.	
  (21,	
  22)	
  
	
  
a	
  

	
  
	
  

b
	
  

	
  

c	
  

d	
  

	
  
	
  

e	
  

f	
  

	
  
	
  
Figure	
  I-­‐6:	
   	
   a.	
   The	
  contoured	
  2Fo-­‐Fc	
   electron	
   density	
   map	
   at	
   1.0	
   σ	
  for	
   rearranged	
  
retinol	
   from	
   data	
   collected	
   at	
   11	
   KeV/the	
   highest	
   flux	
   (4QYN,	
   blue).	
   b.	
   The	
  
	
  
chemical	
  
structure	
   of	
   a	
   possible	
   retinoid	
   derivative	
   consistent	
   with	
   the	
  
crystallographic	
  data.	
  c.	
  The	
  overlay	
  of	
  bonafide	
  retinol	
  structure	
  (chain	
  C)	
  from	
  
data	
  
collected	
   with	
   an	
  attenuated	
   X-­‐ray	
  beam	
   at	
   an	
   energy	
   of	
   7	
   KeV	
   (4QZT,	
   Pink)	
  
	
  
and	
   the	
  structure	
  of	
   rearraged	
  retinoid	
  derivative	
   (chain	
  B)	
  from	
  data	
   collected	
  at	
  
11	
  KeV	
  and	
  the	
  highest	
  flux	
  (4QYN,	
  blue).	
  d.	
  The	
  structures	
  of	
  the	
  noncanonical	
  
	
  
retinoid	
   (refined	
   in	
   two	
   conformations)	
   previously	
   seen	
   in	
   the	
   human	
   CRBPII	
  
(2RCT,	
  yellow),	
  and	
  our	
  current	
  rearranged	
   structure	
  (4QYN,	
  blue)	
   were	
  overlaid.	
  
	
   The	
   structures	
   of	
   the	
   canonical	
   and	
   noncanonical	
   retinoid	
   previously	
   seen	
   in	
  
e.	
  
the	
  zebrafish	
  CRBP	
  (1KQW,	
  green),	
  and	
  our	
  current	
  rearranged	
  structure	
  (4QYN,	
  
blue)	
  
were	
   overlaid.	
  f.	
  The	
  overlaid	
   structures	
  of	
  the	
   canonical	
  retinol	
  obtained	
  in	
  
	
  
the	
  rat	
  CRBPII	
  (1OPB,	
  purple)	
  and	
  our	
  current	
  bonafide	
  all-­‐trans-­‐retinol	
  structure	
  
in	
  human	
  CRBPII	
  (4QZT,	
  pink).	
  
	
  
This	
   other	
   retinoid	
   species,	
   postulated	
   to	
   be	
   a	
   degradation	
   product	
   from	
   the	
   original	
  
retinol,	
   is	
   characterized	
   by	
   a	
   torsion	
   angle	
   about	
   the	
   C13-­‐C14	
   bond	
   that	
   is	
   around	
   60°,	
  
inconsistent	
   with	
   the	
   trans	
   double	
   bond	
   between	
   these	
   two	
   carbons	
   expected	
   for	
   retinol.	
  
(Figure	
   I-­‐6b,	
   9a	
   and	
   9b)	
   The	
   hydroxyl	
   group	
   of	
   this	
   species	
   makes	
   hydrogen	
   bonds	
   to	
   both	
  
9	
  

Gln108	
  and	
  Lys40,	
  essentially	
  occupying	
  the	
  same	
  position	
  as	
  the	
  water	
  molecule	
  responsible	
  
for	
   the	
   water	
   mediated	
   interaction	
   seen	
   in	
   our	
   hCRBPII-­‐retinal	
   complex.	
   In	
   rat	
   CRBPII-­‐retinol	
  
complex	
   shows	
   only	
   retinol	
   in	
   the	
   binding	
   pocket,	
   with	
   the	
   hydroxyl	
   group	
   in	
   a	
   distinctly	
  
different	
  position	
  to	
  that	
  seen	
  in	
  the	
  other	
  structures,	
  allowing	
  it	
  to	
  make	
  hydrogen	
  bonds	
  to	
  
both	
  the	
  amide	
  oxygen	
  and	
  nitrogen	
  atoms	
  of	
  Gln108.	
  (Figure	
  I-­‐7c	
  and	
  9c)	
  Since	
  we	
  were	
  not	
  
confident	
   on	
   the	
   previous	
   hCRBPII-­‐retinol	
   complexes,	
   the	
   structure	
   of	
   the	
   hCRBPII-­‐retinol	
  
complex	
  was	
  re-­‐determined	
  in	
  a	
  new	
  P1	
  crystal	
  form	
  with	
  2	
  molecules	
  per	
  asymmetric	
  unit	
  in	
  
our	
  group.	
  	
  (Table	
  II-­‐1)	
  
	
  

b	
  

a	
  
	
  
	
  
	
  
	
  
c	
  

(a)

	
  

Asn91

(b)
Thr1

Gln108

Gln4
W

	
  

W
retinol
Asp91
Val110

	
  

	
  

Gln4

Gln108
retinol

	
   I-­‐7:	
   	
   a.	
   The	
   rearranged	
  
Figure	
  
retinol	
   from	
   data	
  
collected	
   at	
   11	
   (d)
KeV	
   and	
   the	
   highest	
  
flux	
  
(c)
Phe4
Phe4
Thr51
(4QYN,	
  blue)	
  and	
  the	
  interaction	
  of	
  alcohol	
  moiety	
  with	
  the	
  neighboring	
  residues	
  of	
  Gln108	
  
Ile51
and	
  	
  Lys40.	
   b.	
   The	
   structures	
   of	
   the	
   retinal	
   bound	
   hCRBPII	
   (4QYP,	
   cyan,	
   chain	
   B),	
   and	
   our	
  
current	
   rearranged	
   retinol	
   structure	
   (4QYN,	
   blue)	
   were	
   overlaid.	
   The	
   ordered	
   water	
  
Asp91
molecule	
  (W)	
  seen	
   in	
  all-­‐trans-­‐retinal	
  crystal	
  structure,	
  is	
  occupied	
  by	
  the	
  hydroxyl	
  group	
   of	
  
the	
  noncanonical	
  retinoid	
  (indicated	
  in	
  dashed	
  circle).	
  Similar	
  case	
  in	
  retinol	
  structure	
  from	
  
data	
   collected	
   at	
   7	
   KeV	
   and	
   an	
   attenuated	
   beam	
  
Gln108 has	
   been	
   seen.	
   c.	
   Interactions	
   around	
  
Gln108
retinol and retinoid
retinol
Gln108	
  of	
  retinol	
  bound	
  rat	
  CRBPII	
  (1OPB,	
  purple).	
  
10	
  

Asp91

Interestingly,	
  The	
  same	
  retinoid	
  derivative	
  was	
  seen	
  in	
  our	
  structure.	
  (Figure	
   I-­‐6a,	
   6d	
   and	
   6e)	
  
The	
  hydroxyl	
  of	
  the	
  retinoid	
  derivative	
  makes	
  hydrogen	
  bonds	
  to	
  both	
  Gln108	
  and	
  Lys40,	
  and	
  
occupies	
   a	
   position	
   that	
   is	
   essentially	
   identical	
   to	
   that	
   of	
   the	
   water	
   in	
   the	
   retinal	
   complex	
  
(Figure	
  I-­‐7).	
  No	
  evidence	
  for	
  retinol	
  is	
  seen	
  in	
  our	
  structures.	
  Though	
  the	
  exact	
  identity	
  of	
  the	
  
retinoid	
  is	
  unclear	
  (Figure	
  I-­‐6b).	
  NMR	
  spectrum	
  of	
  the	
  retinol	
  sample	
  was	
  taken	
  by	
  Dr.	
  Yapici	
  to	
  
determine	
  the	
  purity	
  of	
  the	
  retinol.	
  	
  The	
  NMR	
  showed	
  the	
  retinol	
  to	
  be	
  quite	
  pure	
  with	
  
no	
  significant	
  extraneous	
  peaks.	
  	
  Further,	
  Dr.	
  Yapici	
  performed	
  a	
  series	
  of	
  analytical	
  HPLC	
  of	
  the	
  
sample,	
   which	
   they	
   also	
   showed	
   only	
   a	
   single	
   peak	
   consistent	
   with	
   retinol.	
   (Figure	
   I-­‐8).	
   Since	
  
our	
  crystals	
  were	
  all	
  grown	
  at	
  a	
  relatively	
  low	
  pH	
  of	
  4.6,	
  we	
  investigated	
  the	
  possibility	
  that	
  the	
  
low	
   pH	
   of	
   the	
   crystallization	
   condition	
   could	
   lead	
   to	
   the	
   retinoid	
   derivative	
   by	
   incubating	
  
retinol,	
   both	
   alone	
   and	
   in	
   the	
   presence	
   of	
   hCRBPII	
   overnight.	
   In	
   both	
   cases	
   a	
   decomposition	
  
product	
  was	
  identified	
  by	
  HPLC.	
  	
  However,	
  this	
  product	
  was	
  less	
  polar	
  as	
  evident	
  from	
  the	
  HPLC	
  
retention	
   times,	
   and	
   is	
   inconsistent	
   with	
   a	
   hydroxyl-­‐group	
   containing	
   compound.	
   Then	
   we	
  
considered	
  that	
  the	
  low	
  pH	
  product	
  was	
  a	
  dehydration	
  product,	
  similar	
  to	
  that	
  previously	
  seen,	
  
but	
   was	
   not	
   the	
   compound	
   seen	
   in	
   the	
   structure	
   (Figure	
   I-­‐a,	
   6b).(26)	
   We	
   then	
   grew	
   over	
   30	
  
crystals	
  of	
  the	
  complex,	
  and	
  Dr.	
  Yapici	
  dissolved	
  them	
  in	
  phosphate	
  buffered	
  saline	
  extracted	
  
the	
  	
  retinoids	
  by	
  hexane	
  and	
  analysed	
  them	
  by	
  HPLC	
  and	
  UV/Vis	
  spectroscopy.	
  	
  
	
  all	
  of	
  the	
  other	
  data	
  sets,	
  the	
  human,	
  zebrafish,	
  and	
  both	
  of	
  our	
  hCRBPII	
  complex	
  structures,	
  
were	
   collected	
   at	
   synchrotron	
   sources	
   using	
   about	
   1Å	
   (11	
   KeV)	
   wavelength	
   radiation.	
   (Figure	
  I-­‐
6,	
  9)(23)	
  
	
  

11	
  

	
   	
  
a

b
	
  

.	
  

c

d
.

	
  

e
	
  

Figure	
   I-­‐8:	
   This	
   experiment	
   performed	
   by	
   Dr.	
   Yapici	
   a.	
   UV-­‐vis	
   spectrum	
   of	
   all-­‐trans-­‐
retinol.	
   	
   b.	
   	
   The	
   HPLC	
   trace	
   of	
   all-­‐trans-­‐retinol.	
   c.	
   	
   UV-­‐vis	
   spectrum	
   of	
   the	
   retinol	
  
extracted	
   from	
   the	
   crystals	
   and	
   purified	
   by	
   HPLC.	
  d.	
   HPLC	
   chromatogram	
   of	
   the	
   retinol	
  
extracted	
   from	
   crystals.	
   e.	
   	
   HPLC	
   chromatogram	
   of	
   bonafide	
   all-­‐trans-­‐retinol	
   and	
   the	
  
retinol	
  extracted	
  from	
  crystals.	
  Showing	
  that	
  they	
  are	
  the	
  same	
  species.	
  

This	
  showed	
  only	
  a	
  single	
  peak	
  that	
  co-­‐eluted	
  precisely	
  with	
  retinol,	
  indicating	
  that	
  the	
  
crystals	
   do	
   in	
   fact	
   contain	
   only	
   retinol.	
   (Figure	
   I-­‐8)	
   It	
   therefore	
   appears	
   that	
   the	
   retinoid	
  
derivative	
   seen	
   in	
   these	
   structures	
   is	
   due	
   to	
   X-­‐ray	
   induced	
   damage	
   during	
   data	
   collection.	
  
Consistent	
  with	
  this	
  possibility,	
  It	
  was	
  observed	
  that	
  the	
  yellow	
  colour	
  of	
  the	
  crystal	
  changed	
  to	
  
dark	
   orange	
   after	
   its	
   exposure	
   to	
   the	
   X-­‐ray	
   beam	
   and	
   this	
   colour	
   is	
   maintained	
   to	
   the	
   end	
   of	
  

12	
  

data	
   collection.(27-­‐29)	
   It	
   is	
   interesting	
   to	
   note	
   that	
   for	
   the	
   rat	
   CRBPII	
   retinol	
   complex,	
   the	
   only	
  
bonafide	
  retinol-­‐bound	
  CRBPII	
  structure,	
  data	
  was	
  collected	
  using	
  a	
  Cu	
  Kι	
  home	
  X-­‐ray	
  source,	
  	
  
In	
  an	
  effort	
  to	
  finally	
  produce	
  a	
  bonafide	
  hCRBPII	
  retinol	
  complex,	
  we	
  collected	
  a	
  variety	
  
of	
   data	
   sets	
   at	
   the	
   synchrotron.	
   	
   Since	
   most	
   X-­‐ray	
   damage	
   of	
   protein	
   crystals	
   are	
   thought	
   to	
   be	
  
dosage	
   related.	
   We	
   first	
   collected	
   data	
   at	
   11	
   KeV	
   but	
   with	
   the	
   beam	
   attenuated	
   to	
   less	
   than	
  
20%	
  intensity.	
  The	
  electron	
  density	
  was	
  still	
  consistent	
  with	
  only	
  the	
  rearranged	
  retinol	
  in	
  the	
  
active	
   site	
   (data	
   not	
   shown).	
   Then	
   we	
   investigated	
   the	
   effect	
   of	
   X-­‐ray	
   energy	
   by	
   collecting	
   data	
  
at	
  7	
  KeV,	
  8	
  KeV	
  and	
  9	
  KeV,	
  all	
  using	
  a	
  beam	
  attenuated	
  to	
  20%	
  of	
  the	
  original,	
  and	
  an	
  additional	
  
data	
   set	
   at	
   11	
   KeV.	
   Dr	
   Spencer	
   Anderson	
   from	
   Argonne	
   National	
   Laboratory	
   helped	
   us	
   in	
  
collecting	
   the	
   attenuated	
   data	
   at	
   multiple	
   wavelengths.	
   Surprisingly,	
   the	
   retinoid	
   electron	
  
density	
   in	
   the	
   7	
   KeV	
   data	
   set	
   clearly	
   showed	
   an	
   un-­‐rearranged	
   all-­‐trans-­‐retinol	
   (Figure	
  I-­‐9c).	
  All	
  
the	
   other	
   data	
   sets	
   showed	
   some	
   mixture	
   of	
   probably	
   retinol	
   and	
   the	
   rearranged	
   product.	
  
Together,	
   these	
   data	
   show	
   that	
   the	
   rearranged	
   product	
   is	
   a	
   result	
   of	
   both	
   dosage	
   and	
  
wavelength-­‐dependent	
   X-­‐ray	
   damage,	
   and	
   that	
   data	
   collection	
   at	
   lower	
   energy	
   results	
   in	
   the	
  
first	
  structure	
  of	
  the	
  bonafide	
  hCRBPII-­‐retinol	
  complex.	
  (Figure	
  I-­‐9d,	
  10,	
  1b)(6)	
  

I-­‐5	
  The	
  structural	
  difference	
  between	
  retinol-­‐bound	
  hCRBPII	
  and	
  retinal-­‐bound.	
  	
  
	
  

The	
  difference	
  in	
  binding	
  of	
  retinal	
  versus	
  retinol	
  lies	
  in	
  the	
  orientation	
  of	
  the	
  hydroxyl	
  

group	
   (Figure	
   I-­‐10c,	
   10d).	
   In	
   the	
   hCRBPII	
   retinol	
   complex	
   the	
   hydroxyl	
   is	
   positioned	
   to	
   make	
  
hydrogen	
   bonds	
   with	
   Lys40	
   and	
   Gln108,	
   making	
   hydrogen	
   bonds	
   with	
   both	
   carbonyl	
   oxygen	
  
and	
   nitrogen	
   atoms	
   of	
   Gln108	
   simultaneously	
   (an	
   interaction	
   that	
   would	
   be	
   impossible	
   for	
  
retinal,	
  which	
  cannot	
  act	
  as	
  hydrogen	
  bond	
  donor	
  to	
  the	
  amide	
  carbonyl	
  oxygen).	
  This	
  results	
  in	
  

13	
  

a	
  different	
  orientation	
  around	
  C15,	
  making	
  room	
  for	
  the	
  water	
  molecule	
  between	
  Gln108	
  and	
  
Lys40	
  and	
  the	
  creation	
  of	
  the	
  hydrogen	
  bond	
  to	
  this	
  water	
  molecule.	
  
all-trans-retinol

retinoid moiety

ψ = C15-C14 / C13-C12

ψ = C15-C14 / C13-C12

	
  

a	
  (a)
ψ1 = 174.4˚

C13

	
  

C13

C12
C14

C12
C14

C13

C15
C15

	
  

ψ2 = 127.6˚
retinoid moiety

all-trans-retinol

ψ = C15-C14 / C13-C12

b	
  (b)

ψ = C15-C14 / C13-C12

ψ3 = 157.0˚

	
  

C13

ψ4 = 90.1˚

C12

C12
C14

C13

	
  

C14

C13

C15

C15

c	
  

all-trans-retinol

	
  

ψ = C15-C14 / C13-C12

(c)

	
  

	
  

ψ5 = 178.3˚
C13
C12
C15

	
  
	
  

C14
C13

d	
  

all-trans-retinol

ψ = C15-C14 / C13-C12

(d)

ψ6 = 171.6˚

	
  
C13

	
  

C15

C15

C12
C13

C14
C15

Figure	
   I-­‐9:	
   A	
   comparison	
   of	
   the	
   retinol	
   and	
   retinoid	
   moieties	
   found	
   in	
   the	
   crystal	
  
structures	
   of	
   human,	
   zebra	
   fish	
   and	
   rat	
   CRBPII.	
   a.	
   The	
   retinol	
   (left)	
   and	
   retinoid	
   (right)	
  
moieties	
   found	
   in	
   the	
   previously	
   published	
   hCRBPII	
   (2RCT,	
   yellow).	
   The	
   value	
   of	
   128°	
  
torsion	
  angle	
  (ψ)	
  about	
  the	
  C13-­‐C14	
  bond	
  is	
  inconsistent	
  with	
  a	
  double	
  bond	
  (180°).	
  b.	
  	
  
The	
   retinol	
  and	
   retinoid	
  moieties	
  seen	
  in	
   the	
  crystal	
  structure	
  of	
  zebrafish	
  CRBP	
  (1KQW,	
  
green).	
   	
   The	
   torsion	
   angle	
   about	
   the	
   C13-­‐C14	
   bond	
   is	
   also	
   inconsistent	
   with	
   a	
   double	
  
bond.	
  c.	
  The	
  retinol	
  structure	
  in	
  the	
  rat	
  CRBPII	
  (1OPB,	
  purple).	
  d.	
  	
  The	
  retinol	
  found	
  in	
  the	
  
X-­‐ray	
  structure	
  of	
  hCRBPII	
  using	
  data	
  collected	
  at	
  an	
  energy	
  of	
  7	
  KeV	
  and	
  an	
  attenuated	
  
beam	
  (4QZT,	
  pink)	
  with	
  an	
  approximate	
  180°	
  torsion	
  angle	
  around	
  C13-­‐C14	
  bond.	
  

14	
  

	
  
	
  
	
  

By	
   overlaying	
   these	
   two	
   structures,	
   it	
   is	
   clear	
   that	
   the	
   position	
   of	
   the	
   retinol	
   hydroxyl	
  
group	
   in	
   hCRBPII	
   would	
   in	
   fact	
   clash	
   with	
   this	
   water	
   molecule.	
   The	
   difference	
   in	
   affinity	
   for	
  
retinol	
  versus	
  retinal	
  may	
  be	
  explained	
  by	
  their	
  different	
  interactions	
  with	
  hCRBPII.	
  	
  However,	
  
the	
  physiological	
  relevance	
  of	
  this	
  difference	
  in	
  binding	
  constant	
  is	
  unclear.	
  	
  
	
  
	
  
(a)

a	
  

Gln108

	
  

3.3 Å

b(b)
.	
  

3.6 Å

	
  

2.3 Å
Lys40

	
  

(c)

c	
  

d
.	
  

(d)
2.4 Å
3.6 Å

	
  

Gln108

Gln108

W

W
2.5 Å

	
  

2.5 Å
Lys40

Lys40

Figure	
   I-­‐10:	
  Retinal	
  versus	
  retinol	
  binding	
  in	
  hCRBPII.	
  a.	
  The	
  structure	
  of	
  the	
  retinol-­‐
bound	
  WT	
  hCRBPII.	
   b.	
  The	
  2Fo-­‐Fc	
  electron	
  density	
  map	
  contoured	
  at	
  1.0σ	
  around	
  
retinol	
  of	
   7Kev	
   data.	
  c.	
  The	
  structure	
  of	
   all-­‐trans-­‐retinal	
   bound	
   to	
   WT	
  hCRBPII.	
   d.	
  
The	
   structures	
  of	
  all-­‐trans-­‐retinal	
  (green)	
  and	
  retinol	
  (magenta).	
  The	
   position	
  of	
  an	
  
ordered	
   water	
  molecule	
   (W)	
  occupied	
  by	
  the	
   hydroxyl	
  group	
  of	
  retinol	
  is	
  indicated	
  
in	
  dashed	
  circle.	
  

I-­‐6	
  Ligand	
  Binding	
  in	
  CRBPI	
  versus	
  CRBPII	
  
In	
  rat,	
  CRBPI	
  and	
  II	
  have	
  similar	
  binding	
  affinities	
  for	
  all-­‐trans-­‐retinol.	
   (20)	
  An	
  overlay	
  of	
  
the	
  crystal	
  structures	
  of	
  retinal-­‐bound	
  and	
  retinol-­‐bound	
  hCRBPII	
  and	
  the	
  structure	
  of	
  hCRBPI	
  
(PDB	
  entry	
  1KGL)	
  and	
  rat	
  CRBPI	
  (PDB	
  entry	
  1CRB)	
  illustrate	
  that	
  in	
  hCRBPII	
  the	
  ligand	
  is	
  more	
  
than	
  1Å	
  deeper	
  in	
  the	
  binding	
  pocket.(14)	
  The	
  identity	
  of	
  the	
  amino	
  acid	
  at	
  position	
  51	
  appears	
  

15	
  

to	
   be	
   the	
   deciding	
   factor	
   for	
   ligand	
   position	
   (Figure	
   I-­‐11).	
   Position	
   51	
   is	
   Isoleucine	
   in	
   CRBPI	
  
(both	
  rat	
  and	
  human)	
  while	
  it	
  is	
  threonine	
  in	
  hCRBPII.	
  By	
  comparing	
  the	
  structures	
  of	
  human	
  
and	
  rat	
  CRBPI	
  (PDB	
  entry	
  1CRB)(30),	
  It	
  can	
  be	
  seen	
  that	
  retinal	
  cannot	
  bind	
  hCRBPI	
  in	
  the	
  same	
  
way	
  that	
  it	
  does	
  in	
  hCRBPII.	
  This	
  is	
  due	
  to	
  a	
  steric	
  clash	
  between	
  the	
  side	
  chain	
  of	
  isoleucine	
  and	
  
the	
  retinal	
  carbonyl	
  group.	
  	
  
	
  
	
  

Gln108

Asp91

	
  
	
  
	
  

Thr1

Retinal
Ile51
Lys40

Gln4
Phe4

Thr51

	
  
Figure	
   I-­‐11:	
   The	
  structures	
   of	
  retinol-­‐bound	
   rat	
   CRBPI	
   (magenta	
  carbons	
   and	
  hydrogen	
  
bonds)	
   and	
   retinal-­‐bound	
   hCRBPII	
   (green	
   carbons	
   and	
   hydrogen-­‐bonds)	
   are	
   overlayed.	
  
Amino	
  acid	
   labels	
   are	
  colored	
  similarly	
  when	
   different	
   in	
   the	
  two	
  structures.	
  	
   Note	
  the	
  
change	
  in	
  the	
  water	
  network	
  introduced	
  by	
  Phe3.	
  	
  	
  	
  
	
  

The	
  combination	
  of	
  the	
  addition	
  of	
  the	
  larger	
  isoleucine	
  at	
  position	
  51,	
  and	
  the	
  altered	
  

position	
  of	
  the	
  ligand	
  leave	
  no	
  room	
  for	
  a	
  water	
  molecule	
  between	
  Gln108	
  and	
  Lys40.	
  In	
  both	
  
CRBPI	
   structures	
   a	
   hydrogen	
   bond	
   is	
   formed	
   between	
   Gln108	
   and	
   the	
   retinol	
   hydroxyl	
   group	
  
(Figure	
   I-­‐11).	
   Though	
   there	
   is	
   no	
   structure	
   of	
   a	
   CRBPI	
   bound	
   to	
   retinal,	
   it	
   is	
   clear	
   that	
   a	
  
hydrogen	
  bond	
  to	
  Gln108	
  can	
  only	
  be	
  made	
  if	
  the	
  amide	
  nitrogen	
  is	
  pointing	
  toward	
  the	
  retinal	
  
carbonyl,	
  because	
  it	
  must	
  be	
  the	
  hydrogen	
  donor.	
  	
  
This	
  is	
  opposite	
  to	
  the	
  orientation	
  of	
  Gln108	
  in	
  hCRBPII,	
  where	
  the	
  orientation	
  of	
  Gln108	
  
is	
  determined	
  by	
  Gln4	
  and	
  a	
  water	
  network	
  as	
  described	
  previously.	
  However,	
  in	
  CRBPI,	
  there	
  is	
  

16	
  

a	
  phenylalanine	
  at	
  position	
  4	
  (Phe3	
  in	
  hCRBPI)	
  instead	
  of	
  glutamine,	
  which	
  may	
  allow	
  Gln108	
  to	
  
rotate	
  its	
  sidechain,	
  such	
  that	
  the	
  amide	
  nitrogen	
  is	
  positioned	
  to	
  make	
  a	
  hydrogen	
  bond	
  with	
  
the	
  aldehyde	
  carbonyl,	
  allowing	
  retinal	
  to	
  bind	
  CRBPI	
  very	
  similarly	
  to	
  that	
  of	
  retinol,	
  resulting	
  
in	
  the	
  similar	
  affinity	
  seen	
  for	
  the	
  two	
  ligands	
  in	
  CRBPII.	
  	
  
In	
   an	
   attempt	
   to	
   understand	
   the	
   role	
   of	
   the	
   residue	
   at	
   position	
   51,	
   the	
   hCRBPII-­‐T51I	
  
mutant	
   was	
   produced	
   and	
   its	
   binding	
   affinity	
   for	
   retinol	
   and	
   retinal	
   measured.	
   The	
   dissociation	
  
constants	
   of	
   this	
   mutant	
   for	
   both	
   all-­‐trans-­‐retinol	
   and	
   all-­‐trans-­‐retinal	
   were	
   similar	
   to	
   that	
   of	
  
the	
   wild	
   type	
   protein	
   (Figure	
   I-­‐3b,	
   3d)	
   indicating	
   that	
   position	
   51	
   is	
   not	
   responsible	
   for	
   the	
  
differences	
  in	
  affinity	
  in	
  the	
  two	
  proteins.	
  Unfortunately	
  attempts	
  at	
  crystallization	
  of	
  hCRBPII-­‐
T51I	
   bound	
   to	
   all-­‐trans-­‐retinal	
   were	
   not	
   successful.	
   However,	
   crystals	
   of	
   the	
   retinol-­‐bound	
  
CRBPII-­‐T51V	
   complex	
   were	
   obtained	
   and	
   its	
   structure	
   determined.	
   	
   However,	
   this	
   structure,	
  
collected	
   with	
   an	
   intense	
   X-­‐ray	
   beam	
   with	
   an	
   energy	
   of	
   11	
   KeV,	
   also	
   showed	
   the	
   rearranged	
  
retinol	
   product	
   in	
   the	
   active	
   site,	
   making	
   it	
   difficult	
   to	
   determine	
   the	
   effect	
   of	
   a	
   hydrophobic	
  
residue	
  at	
  position	
  51	
  on	
  bonafide	
  retinol	
  binding.	
  

I-­‐7	
  Conclusion	
  
Based	
  on	
  the	
  crystal	
  structure	
  of	
  holo	
  wt	
  hCRBPII	
  with	
  all-­‐trans-­‐retinal,	
  there	
  are	
  at	
  least	
  
three	
  distinct	
  binding	
  modes	
  for	
  retinoids	
  in	
  CRBPI	
  and	
  II.	
  In	
  the	
  human	
  and	
  rat	
  CRBPII	
  retinol-­‐
bound	
   structures,	
   retinol	
   is	
   able	
   to	
   make	
   hydrogen	
   bonds	
   with	
   both	
   the	
   carbonyl	
   and	
   amide	
  
nitrogen	
  of	
  the	
  side	
  chain	
  of	
  Gln108.	
  This	
  mode	
  of	
  interaction	
  is	
  not	
  feasible	
  in	
  retinal,	
  due	
  to	
  
the	
  lack	
  of	
  a	
  hydrogen	
  bond	
  donor,	
  resulting	
  in	
  a	
  conformational	
  change	
  of	
  the	
  carbonyl	
  leading	
  
to	
   a	
   water-­‐mediated	
   interaction	
   between	
   the	
   retinal	
   carbonyl	
   and	
   Lys40	
   and	
   Gln108.	
   This	
   is	
  
consistent	
   with	
   the	
   higher	
   binding	
   affinity	
   for	
   retinol	
   relative	
   to	
   retinal	
   in	
   hCRBPII.	
   Since	
   retinol	
  
17	
  

is	
  able	
  to	
  make	
  three	
  hydrogen	
  bonds	
  to	
  Gln108	
  and	
  Lys40.	
  Also,	
  the	
  presence	
  of	
  Ile51	
  causes	
  a	
  
translation	
  of	
  the	
  ligand	
  in	
  the	
  CRBPI	
  binding	
  pocket,	
  resulting	
  in	
  only	
  a	
  single	
  hydrogen	
  bond	
  
being	
  made	
  between	
  Gln108	
  and	
  the	
  retinol	
  hydroxyl	
  group.	
  

I-­‐8	
  Experimental	
  
I-­‐8-­‐1	
  Material	
  and	
  Method:	
  Protein	
  Expression	
  and	
  Purification	
  
The	
  hCRBPII	
  gene	
  was	
  purchased	
  from	
  ATCC	
  and	
  cloned	
  in	
  pET17b	
  vector,	
  with	
  using	
  the	
  
NdeI	
   and	
   XhoI	
   restriction	
   enzyme	
   in	
   the	
   N-­‐	
   and	
   C	
   terminus	
   respectively.	
   1	
   ΟL	
   of	
   the	
   resulting	
  
plasmid	
  were	
  transformed	
  in	
  50	
  ΟL	
  of	
  E.coli	
  DH5ι	
  (NovagenŽ)	
  competent	
  cells.	
  The	
  cells	
  were	
  
incubated	
   for	
   30	
   min	
   in	
   ice	
   and	
   heat	
   shocked	
   at	
   42	
   °C	
   for	
   28	
   seconds,	
   then	
   450	
   ΟL	
   of	
   Luria-­‐
Bertani	
  broth	
  (LB)	
  was	
  added	
  and	
  the	
  cells	
  were	
  incubated	
  at	
  37°C	
  for	
  an	
  hour.	
  The	
  resulting	
  
mixture	
   were	
   spread	
   on	
   an	
   LB	
   agar	
   plate	
   with	
   (ampicillin	
   100	
   Οg/mL)	
   and	
   incubated	
   at	
   37°C	
   for	
  
16	
  hours.	
  A	
  single	
  colony	
  was	
  picked	
  from	
  the	
  plate	
  and	
  inoculated	
  in	
  5	
  mL	
  of	
  LB	
  media	
  contains	
  
100	
   Οg/mL	
   ampicillin.	
   The	
   cell	
   culture	
   was	
   grown	
   over	
   night	
   (12-­‐16	
   hours)	
   at	
   37°C,	
   then	
   the	
  
media	
   were	
   centrifuged	
   at	
   15000	
   rpm	
   for	
   1	
   min.	
   DNA	
   extraction	
   and	
   isolation	
   from	
   the	
   cell	
  
pellet	
  was	
  done	
  according	
  to	
  the	
  manufacturers	
  instructions	
  (Promega	
  Wizard	
  and	
  SV	
  Miniprep	
  
(A1330)	
  DNA	
  purification	
  kit).	
  The	
  target	
  gene	
  was	
  transformed	
  into	
  BL21(DE3)	
  (InvitrogenTM)	
  
E.coli	
  competent	
  cells	
  for	
  protein	
  expression.	
  A	
  single	
  colony	
  was	
  picked	
  and	
  inoculated	
  in	
  25	
  
mL	
   of	
   LB,	
   containing	
   100	
   Οg/mL	
   ampicillin	
   at	
   37°C	
   for	
   12-­‐16	
   hours.	
   This	
   media	
   was	
   then	
  
transferred	
   into	
   1L	
   of	
   LB	
   media	
   with	
   ampicillin	
   (100	
   mg/L)	
   to	
   grow	
   large	
   scale	
   and	
   incubated	
   at	
  
37°C	
  until	
  OD600	
  reached	
  0.5-­‐1.0.	
  The	
  protein	
  expression	
  was	
  induced	
  with	
  1	
  mM	
  isopropyl	
  β-­‐D-­‐
1	
   thiogalactopyranoside	
   (IPTG)	
   (purchased	
   from	
   Gold	
   Biotechnology)	
   overnight	
   at	
   25°C.	
   The	
  

18	
  

cells	
   were	
   harvested	
   by	
   centrifugation	
   at	
   5000	
   rpm	
   for	
   20	
   min.	
   The	
   harvested	
   cells	
   were	
  
resuspended	
  in	
  Tris	
  buffer	
  (10	
  mM	
  Tris,	
  10	
  mM	
  NaCl	
  pH	
  8.0,	
  50mL).	
  The	
  suspended	
  cells	
  were	
  
lysed	
  by	
  sonication	
  and	
  the	
  lysed	
  cells	
  were	
  centrifuged	
  at	
  4°C	
  (14,000	
  rpm,	
  20	
  min).	
  	
  
The	
   supernatant	
   was	
   loaded	
   on	
   a	
   Fast	
   Q	
   anion	
   exchange	
   resin	
   (purchased	
   from	
   GE	
  
healthcare),	
  which	
  was	
  equilibrated	
  with	
  buffer	
  A	
  (10	
  mM	
  Tris,	
  10	
  mM	
  NaCl	
  pH	
  8.0).	
  Then	
  after	
  
the	
  binding,	
  the	
  resin	
  was	
  washed	
  three	
  times	
  with	
  50	
  mL	
  of	
  buffer	
  A.	
  The	
  protein	
  was	
  eluted	
  
with	
   40	
   mL	
   of	
   elution	
   buffer	
   (10	
   mM	
   Tris,	
   100	
   mM	
   NaCl	
   pH	
   8.0).	
   The	
   eluted	
   protein	
   was	
  
desalted	
   using	
   a	
   CentriprepŽ	
   centrifugal	
   filter	
   (10kDa	
   cutoff)	
   at	
   2500	
   rpm.	
   Protein	
   was	
   then	
  
purified	
  on	
  a	
  15Q	
  anion	
  exchange	
  column	
  using	
  a	
  BioLogic	
  DuoFlow	
  system	
  for	
  the	
  second	
  step	
  
of	
  purification.	
  The	
  purity	
  of	
  the	
  protein	
  was	
  determined	
  with	
  SDS-­‐PAGE.	
  	
  

I-­‐8-­‐2	
  	
  Kd	
  Determination	
  via	
  Fluorescence	
  Quenching	
  Assay	
  
The	
   dissociation	
   constant	
   measurement,	
   (Kd)	
   for	
   both	
   all-­‐trans-­‐retinol	
   and	
   all-­‐trans-­‐
retinal	
   (purchased	
   from	
   Sigma)	
   with	
   hCRBPII	
   was	
   determined	
   by	
   fluorescence	
   quenching	
   assay.	
  
All	
   samples	
   were	
   stored	
   in	
   glass	
   containers,	
   as	
   many	
   plastic	
   containers	
   will	
   leach	
   fluorescent	
  
impurities	
  into	
  the	
  sample.	
  Salinized	
  glassware	
  was	
  used	
  to	
  avoid	
  loss	
  of	
  protein	
  and	
  a	
  change	
  
in	
  protein	
  concentration.	
  	
  
The	
   cuvette	
   was	
   allowed	
   to	
   sit	
   with	
   3	
   mL	
   of	
   a	
   0.01%	
   gelatin	
   containing	
   PBS	
   buffer	
   (4	
   mM	
  
NaH2PO4,	
   16	
   mM	
   Na2HPO4,	
   150	
   mM	
   NaCl,	
   pH	
   7.3)	
   for	
   30	
   minutes.	
   The	
   clean	
   cuvette	
   was	
  
rinsed	
  once	
  with	
  distilled	
  water,	
  and	
  the	
  protein	
  solution	
  was	
  added	
  (3	
  mL,	
  20	
  ΟM).	
  The	
  sample	
  
was	
  excited	
  at	
  280	
  nm	
  with	
  an	
  excitation	
  slit	
  width	
  of	
  1.5	
  nm.	
  The	
  fluorescence	
  was	
  measured	
  
at	
  the	
  peak	
  maximum	
  (345	
  nm).	
  This	
  was	
  repeated	
  three	
  times	
  until	
  a	
  stable	
  emission	
  intensity	
  
at	
  350	
  nm	
  was	
  reached.	
  Retinal	
  was	
  added	
  to	
  the	
  cuvette	
  in	
  varying	
  amounts	
  from	
  a	
  1.5	
  mM	
  
19	
  

stock	
   solution	
   in	
   ethanol	
   maintained	
   in	
   the	
   dark.	
   Care	
   was	
   taken	
   to	
   ensure	
   that	
   the	
   EtOH	
  
volume	
   remained	
   below	
   2%.	
   The	
   titration	
   was	
   complete	
   when	
   there	
   was	
   no	
   observable	
  
quenching	
   of	
   fluorescence	
   upon	
   addition	
   of	
   the	
   chromophore.	
   This	
   was	
   plotted	
   as	
  
concentration	
   of	
   chromophore	
   versus	
   relative	
   fluorescence	
   intensity.	
   The	
   data	
   were	
   analyzed	
  
by	
  nonlinear	
  square	
  fit	
  of	
  the	
  equation.	
  The	
  Kd	
  for	
  each	
  hCRBPII	
  was	
  determined	
  according	
  to	
  
the	
  method	
  previously	
  described.(31)	
  

I-­‐8-­‐3	
  Crystallization	
  and	
  Structure	
  Determination	
  
The	
   pure	
   WT	
   hCRBPII	
   protein	
   was	
   concentrated	
   to	
   5-­‐10	
   mg/mL.	
   The	
   complexes	
   of	
  
protein	
   with	
   all-­‐trans-­‐retinol	
   were	
   prepared	
   by	
   adding	
   3-­‐4	
   equivalents	
   of	
   retinal	
   and	
   retinol	
  
solution	
  (30	
  mM	
  retinol	
  in	
  ethanol).	
  The	
  final	
  concentration	
  of	
  ethanol	
  in	
  protein	
  solution	
  was	
  
kept	
   below	
   10%	
   V/V.	
   The	
   mixtures	
   of	
   protein	
   and	
   ligands	
   were	
   incubated	
   at	
   room	
   temperature	
  
in	
  Black	
  LiteSafeTM	
  Microcentrifuge	
  Tubes	
  for	
  2	
  hours.	
  The	
  crystal	
  of	
  the	
  complex	
  of	
  protein	
  and	
  
ligands	
  were	
  prepared	
  at	
  room	
  temperature	
  by	
  hanging	
  drop	
  vapor	
  diffusion	
  in	
  Limbro	
  plates	
  
wrapped	
  in	
  aluminum	
  foil	
  to	
  prevent	
  light-­‐initiated	
  degradation	
  of	
  the	
  retinal	
  and	
  retinol.	
  The	
  
best	
  crystals	
  were	
  grown	
  using	
  Hampton	
  Research	
  Screens	
  (the	
  best	
  crystals	
  growing	
  in	
  30-­‐35%	
  
PEG	
  4000	
  (Sigma-­‐Aldrich),	
  0.1	
  M	
  sodium	
  acetate	
  (Columbus	
  Chemical	
  Industry)	
  pH	
  4.6-­‐4.8	
  and	
  
0.1	
   M	
   ammonium	
   acetate	
   (J.	
   T.	
   Baker)	
   with	
   1	
   ΟL	
   of	
   hCRBPII-­‐ligand	
   complex	
   and	
   1	
   ΟL	
   of	
   mother	
  
liquor).	
  The	
  crystals	
  appeared	
  after	
  3	
  days	
  and	
  reached	
  their	
  maximum	
  size	
  in	
  one	
  week.	
  The	
  
crystals	
  were	
  flash	
  frozen	
  in	
  liquid	
  nitrogen	
  using	
  a	
  cryo-­‐protectant	
  solution	
  (30%	
  PEG	
  4000,	
  0.1	
  
M	
   sodium	
   acetate	
   pH	
   4.6-­‐4.8,	
   0.1	
   M	
   ammonium	
   acetate,	
   20%	
   glycerol)	
  and	
   stored	
   in	
   a	
   liquid	
  
nitrogen	
  dewar.	
  	
  

20	
  

All	
   of	
   the	
   diffraction	
   data	
   were	
   collected	
   at	
   beamline	
   21-­‐ID-­‐D,	
   LS-­‐CAT	
   (Argonne	
   National	
  
Laboratory,	
   Advanced	
   Photon	
   Source,	
   Chicago,	
   IL)	
   using	
   a	
   MAR300	
   detector	
   and	
   1.00Å	
  
wavelength	
   radiation	
   at	
   100K.	
   The	
   diffraction	
   data	
   were	
   indexed	
   using	
   the	
   HKL2000	
   software	
  
package.(32)	
   The	
   structure	
   was	
   solved	
   by	
   molecular	
   replacement	
   using	
   the	
   MOLREP	
   program	
  
implemented	
   in	
   the	
   ccp4	
   programing	
   package	
   using	
   human	
   CRBPII	
   (Protein	
   data	
   bank,	
  
accession	
  code	
  2RCQ(22))	
  as	
  model.	
  Initial	
  electron	
  density	
  maps	
  and	
  structure	
  refinement	
  was	
  
performed	
  using	
  REFMAC5	
  in	
  the	
  CCP4	
  suite.(33,	
  34)	
   (35)All	
  rebuilding	
  and	
  placement	
  of	
  ordered	
  
water	
   molecules	
   was	
   done	
   manually	
   using	
   COOT.(36)	
   The	
   chromophore	
   was	
   created	
   using	
   j-­‐
ligand	
  and	
  manually	
  fitted	
  in	
  the	
  electron	
  density	
  at	
  the	
  final	
  step	
  of	
  the	
  refinement.	
  	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
21	
  

	
  Table	
  I-­‐1:	
  Data-­‐collection	
  and	
  refinement	
  statistics.	
  	
  
	
  
	
  
	
  
wt	
  hCRBPII-­‐retinol-­‐
wt	
  hCRBPII-­‐retinol	
  
7KeV	
  
	
  
Space	
  
group	
  
P21	
  
P1	
  
a(Å)	
  
	
  

34.65	
  

b(Å)	
  

	
  

75.14	
  

c	
  (Å)	
  

54.65	
  

	
   )	
  
ι(°	
  

90.00	
  

β(°	
  
	
   )	
  

100.78	
  

γ(°	
  
	
   )	
  

90.00	
  

	
  

54.18	
  

	
  

68.44	
  

Total	
  reflection	
  

306503	
  

Unique	
  Reflection	
  

87302	
  

Completeness	
  
(%)	
  
	
  

99.8(84.0) 	
  

Molecules	
  
per	
  
	
  
Asymmetric	
  Unit	
  

2	
  

a

	
  
36.48	
  
	
  
54.14	
  
68.30	
  

107.72	
  

107.64	
  

97.19	
  

96.94	
  

103.59	
  

103.71	
  

a

50.00-­‐1.19(1.21-­‐1.19) 	
  

	
  

	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  P1	
  

36.41	
  

Resolution	
  (Å)	
  

	
  

wt	
  hCRBPII-­‐retinol-­‐	
  	
  
11KeV	
  

29.79-­‐1.89(1.94-­‐1.89) 	
  
76284	
  

34.661-­‐1.496	
  (1.51-­‐
a	
  
1.50)
	
  
358399	
  

34646	
  

76284	
  

a

a

a

92.36(86.0) 	
  

95.20(88.0) 	
  

4	
  

4	
  

a

20.19(2.04) 	
  
Average	
  
I/σ	
  
	
  
	
  
	
  
a
R	
   merge(%)	
  
9.1(43.4) 	
  
	
  
	
  	
  	
  	
  	
  
Mosaicity	
  (°)	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  0.486	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
   	
  

	
  

	
  
	
  a

	
  

a

12.97(6.23) 	
  
a

a

26.45(2.41) 	
  
a

7.8(32.5) 	
  
	
  

4.2(59.8) 	
  
	
  

0.58	
  

0.58	
  

Note:	
  Values	
  in	
  the	
  parenthesis	
  refer	
  to	
  the	
  last	
  resolution	
  shell.	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

22	
  

Table	
  I-­‐1	
  	
  (cont’d)	
  

!
Refinement!
statistics!
Rwork/!Rfree!(%)!

!

	
  
	
  

	
  
	
  

Bond!Length!(Å)!

0.0297!

0.008!

0.007!

Angle!(°)!

2.381!

1.079!

1.066! 	
  

B!average!for!main!
9.148!
34.628!
chain!
B!average!for!
16.410!
45.016!
water!and!side!
	
  
chain!
Number!of!water!
507!
331!
molecules!
!
!
!
!
!
Ramachandran!plot!
97.10!
96.04!
Most!favored!(%)!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
2.07!
3.21!
Allowed!(%)!
0.83!
2.12!
Outliers!(%)!

23.69! 	
  

	
  

!39.549! 	
  
468! 	
  
!
!
	
  
96.20!
2.28! 	
  
2.49!
	
  

	
  

PDB#!

4QYN!

4QZT!

	
  
	
  
	
  
	
  
	
  
	
  
	
  

23	
  

4QZU!

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
REFERENCES	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

24	
  

REFERENCES	
  
	
  
	
  
	
  
1.	
  
	
  
2.	
  

	
  
3.	
  
	
  
4.	
  
	
  
5.	
  

	
  
6.	
  
	
  
7.	
  
	
  
8.	
  
	
  
9.	
  
	
  
10.	
  

	
  

Ong,	
  D.	
  E.	
  (1984)	
  A	
  NOVEL	
  RETINOL-­‐BINDING	
  PROTEIN	
  FROM	
  RAT	
  -­‐	
  PURIFICATION	
  AND	
  
PARTIAL	
  CHARACTERIZATION,	
  Journal	
  of	
  Biological	
  Chemistry	
  259,	
  1476-­‐1482.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  crystallographica.	
  Section	
  
D,	
  Biological	
  crystallography	
  70,	
  3226-­‐3232.	
  
Veerkamp,	
   J.	
   H.,	
   Vankuppevelt,	
   T.,	
   Maatman,	
   R.,	
   and	
   Prinsen,	
   C.	
   F.	
   M.	
   (1993)	
  
STRUCTURAL	
   AND	
   FUNCTIONAL-­‐ASPECTS	
   OF	
   CYTOSOLIC	
   FATTY-­‐ACID-­‐BINDING	
  
PROTEINS,	
  Prostaglandins	
  Leukotrienes	
  and	
  Essential	
  Fatty	
  Acids	
  49,	
  887-­‐906.	
  
Banaszak,	
   L.,	
   Winter,	
   N.,	
   Xu,	
   Z.	
   H.,	
   Bernlohr,	
   D.	
   A.,	
   Cowan,	
   S.,	
   and	
   Jones,	
   T.	
   A.	
   (1994)	
  
LIPID-­‐BINDING	
   PROTEINS	
   -­‐	
   A	
   FAMILY	
   OF	
   FATTY-­‐ACID	
   AND	
   RETINOID	
   TRANSPORT	
  
PROTEINS,	
  Advances	
  in	
  Protein	
  Chemistry,	
  Vol	
  45	
  45,	
  89-­‐151.	
  
Vaezeslami,	
   S.,	
   Mathes,	
   E.,	
   Vasilelou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2006)	
   The	
  
structure	
   of	
   apo-­‐wild-­‐type	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   at	
   1.4	
   A	
   and	
   its	
  
relationship	
   to	
   ligand	
   binding	
   and	
   nuclear	
   translocation,	
   Journal	
   of	
   Molecular	
   Biology	
  
363,	
  687-­‐701.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins-­‐-­‐insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  lipid	
  research	
  43,	
  328-­‐349.	
  
Lucke,	
   C.,	
   Huang,	
   S.,	
   Rademacher,	
   M.,	
   and	
   Ruterjans,	
   H.	
   (2002)	
   New	
   insights	
   into	
  
intracellular	
   lipid	
   binding	
   proteins:	
   The	
   role	
   of	
   buried	
   water,	
   Protein	
   science	
   :	
   a	
  
publication	
  of	
  the	
  Protein	
  Society	
  11,	
  2382-­‐2392.	
  
Schaap,	
   F.	
   G.,	
   van	
   der	
   Vusse,	
   G.	
   J.,	
   and	
   Glatz,	
   J.	
   F.	
   (2002)	
   Evolution	
   of	
   the	
   family	
   of	
  
intracellular	
   lipid	
   binding	
   proteins	
   in	
   vertebrates,	
   Molecular	
   and	
   cellular	
   biochemistry	
  
239,	
  69-­‐77.	
  
Smathers,	
   R.	
   L.,	
   and	
   Petersen,	
   D.	
   R.	
   (2011)	
   The	
   human	
   fatty	
   acid-­‐binding	
   protein	
   family:	
  
evolutionary	
  divergences	
  and	
  functions,	
  Human	
  genomics	
  5,	
  170-­‐191.	
  
Franzoni,	
   L.,	
   Lucke,	
   C.,	
   Perez,	
   C.,	
   Cavazzini,	
   D.,	
   Rademacher,	
   M.,	
   Ludwig,	
   C.,	
   Spisni,	
   A.,	
  
Rossi,	
  G.	
  L.,	
  and	
  Ruterjans,	
  H.	
  (2002)	
  Structure	
  and	
  backbone	
  dynamics	
  of	
  apo-­‐	
  and	
  holo-­‐
cellular	
   retinol-­‐binding	
   protein	
   in	
   solution,	
   Journal	
   of	
   Biological	
   Chemistry	
   277,	
   21983-­‐
21997.	
  

25	
  

11.	
  

	
  
12.	
  

	
  
13.	
  

	
  
14.	
  
	
  
15.	
  
	
  
16.	
  

	
  
17.	
  
	
  
18.	
  
	
  
19.	
  
	
  
20.	
  
	
  

Crow,	
   J.	
   A.,	
   and	
   Ong,	
   D.	
   E.	
   (1985)	
   CELL-­‐SPECIFIC	
   IMMUNOHISTOCHEMICAL	
  
LOCALIZATION	
   OF	
   A	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   (TYPE-­‐2)	
   IN	
   THE	
   SMALL-­‐
INTESTINE	
  OF	
  RAT,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  
of	
  America	
  82,	
  4707-­‐4711.	
  
Vogel,	
  S.,	
  Mendelsohn,	
  C.	
  L.,	
  Mertz,	
  J.	
  R.,	
  Piantedosi,	
  R.,	
  Waldburger,	
  C.,	
  Gottesman,	
  M.	
  
E.,	
  and	
  Blaner,	
  W.	
  S.	
  (2001)	
  Characterization	
  of	
  a	
  new	
  member	
  of	
  the	
  fatty	
  acid-­‐binding	
  
protein	
   family	
   that	
   binds	
   all-­‐trans-­‐retinol,	
   Journal	
   of	
   Biological	
   Chemistry	
   276,	
   1353-­‐
1360.	
  
Folli,	
   C.,	
   Calderone,	
   V.,	
   Ottonello,	
   S.,	
   Bolchi,	
   A.,	
   Zanotti,	
   G.,	
   Stoppini,	
   M.,	
   and	
   Berni,	
   R.	
  
(2001)	
   Identification,	
   retinoid	
   binding,	
   and	
   x-­‐ray	
   analysis	
   of	
   a	
   human	
   retinol-­‐binding	
  
protein,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  98,	
  3710-­‐3715.	
  
Li,	
  E.,	
  Demmer,	
  L.	
  A.,	
  Sweetser,	
  D.	
  A.,	
  Ong,	
  D.	
  E.,	
  and	
  Gordon,	
  J.	
  I.	
  (1986)	
  ISOLATION	
  OF	
  A	
  
CLONED	
   CDNA-­‐ENCODING	
   AN	
   ABUNDANT	
   INTESTINAL	
   RETINOL	
   BINDING-­‐PROTEIN,	
  
Clinical	
  Research	
  34,	
  A685-­‐A685.	
  
Li,	
   E.,	
   and	
   Norris,	
   A.	
   W.	
   (1996)	
   Structure/function	
   of	
   cytoplasmic	
   vitamin	
   A-­‐binding	
  
proteins,	
  Annual	
  Review	
  of	
  Nutrition	
  16,	
  205-­‐234.	
  
Li,	
   E.,	
   Qian,	
   S.	
   J.,	
   Winter,	
   N.	
   S.,	
   Davignon,	
   A.,	
   Levin,	
   M.	
   S.,	
   and	
   Gordon,	
   J.	
   I.	
   (1991)	
  
FLUORINE	
   NUCLEAR-­‐MAGNETIC-­‐RESONANCE	
   ANALYSIS	
   OF	
   THE	
   LIGAND-­‐BINDING	
  
PROPERTIES	
   OF	
   2	
   HOMOLOGOUS	
   RAT	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEINS	
  
EXPRESSED	
  IN	
  ESCHERICHIA-­‐COLI,	
  Journal	
  of	
  Biological	
  Chemistry	
  266,	
  3622-­‐3629.	
  
Calderone,	
   V.,	
   Folli,	
   C.,	
   Marchesani,	
   A.,	
   Berni,	
   R.,	
   and	
   Zanotti,	
   G.	
   (2002)	
   Identification	
  
and	
   structural	
   analysis	
   of	
   a	
   zebrafish	
   apo	
   and	
   holo	
   cellular	
   retinol-­‐binding	
   protein,	
  
Journal	
  of	
  Molecular	
  Biology	
  321,	
  527-­‐535.	
  
Tarter,	
  M.,	
  Capaldi,	
  S.,	
  Carrizo,	
  M.	
  E.,	
  Ambrosi,	
  E.,	
  Perduca,	
  M.,	
  and	
  Monaco,	
  H.	
  L.	
  (2008)	
  
Crystal	
   structure	
   of	
   human	
   cellular	
   retinol-­‐binding	
   protein	
   11	
   to	
   1.2	
   angstrom	
  
resolution,	
  Proteins-­‐Structure	
  Function	
  and	
  Bioinformatics	
  70,	
  1626-­‐1630.	
  
Winter,	
   N.	
   S.,	
   Bratt,	
   J.	
   M.,	
   and	
   Banaszak,	
   L.	
   J.	
   (1993)	
   CRYSTAL-­‐STRUCTURES	
   OF	
   HOLO	
  
AND	
   APO-­‐CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN-­‐II,	
   Journal	
   of	
   Molecular	
   Biology	
   230,	
  
1247-­‐1259.	
  
Kane,	
  M.	
  A.,	
  Bright,	
  F.	
  V.,	
  and	
  Napoli,	
  J.	
  L.	
  (2011)	
  Binding	
  affinities	
  of	
  CRBPI	
  and	
  CRBPII	
  
for	
  9-­‐cis-­‐retinoids,	
  Biochimica	
  Et	
  Biophysica	
  Acta-­‐General	
  Subjects	
  1810,	
  514-­‐518.	
  

26	
  

21.	
  
	
  
22.	
  

	
  
23.	
  
	
  
24.	
  
	
  
25.	
  
	
  
26.	
  

	
  
27.	
  
	
  
28.	
  
	
  
29.	
  
	
  
30.	
  
	
  
31.	
  

Malpeli,	
   G.,	
   Stoppini,	
   M.,	
   Zapponi,	
   M.	
   C.,	
   Folli,	
   C.,	
   and	
   Berni,	
   R.	
   (1995)	
   INTERACTIONS	
  
WITH	
   RETINOL	
   AND	
   RETINOIDS	
   OF	
   BOVINE	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN,	
  
European	
  Journal	
  of	
  Biochemistry	
  229,	
  486-­‐493.	
  
McBee,	
   J.	
   K.,	
   Kuksa,	
   V.,	
   Alvarez,	
   R.,	
   de	
   Lera,	
   A.	
   R.,	
   Prezhdo,	
   O.,	
   Haeseleer,	
   F.,	
   Sokal,	
   I.,	
  
and	
   Palczewski,	
   K.	
   (2000)	
   Isomerization	
   of	
   all-­‐trans-­‐retinol	
   to	
   cis-­‐retinols	
   in	
   bovine	
  
retinal	
   pigment	
   epithelial	
   cells:	
   Dependence	
   on	
   the	
   specificity	
   of	
   retinoid-­‐binding	
  
proteins,	
  Biochemistry	
  39,	
  11370-­‐11380.	
  
Borshchevskiy,	
  V.	
  I.,	
  Round,	
  E.	
  S.,	
  Popov,	
  A.	
  N.,	
  Bueldt,	
  G.,	
  and	
  Gordeliy,	
  V.	
  I.	
  (2011)	
  X-­‐
ray-­‐Radiation-­‐Induced	
   Changes	
   in	
   Bacteriorhodopsin	
   Structure,	
   Journal	
   of	
   Molecular	
  
Biology	
  409,	
  813-­‐825.	
  
Matsui,	
  Y.,	
  Sakai,	
  K.,	
  Murakami,	
  M.,	
  Shiro,	
  Y.,	
  Adachi,	
  S.,	
  Okumura,	
  H.,	
  and	
  Kouyama,	
  T.	
  
(2002)	
  Specific	
  damage	
  induced	
  by	
  X-­‐ray	
  radiation	
  and	
  structural	
  changes	
  in	
  the	
  primary	
  
photoreaction	
  of	
  bacteriorhodopsin,	
  Journal	
  of	
  Molecular	
  Biology	
  324,	
  469-­‐481.	
  
Schreckenbach,	
   T.,	
   Walckhoff,	
   B.,	
   and	
   Oesterhelt,	
   D.	
   (1977)	
   STUDIES	
   ON	
   RETINAL-­‐
PROTEIN	
  INTERACTION	
  IN	
  BACTERIORHODOPSIN,	
  European	
  Journal	
  of	
  Biochemistry	
  76,	
  
499-­‐511.	
  
Cowan,	
   S.	
   W.,	
   Newcomer,	
   M.	
   E.,	
   and	
   Jones,	
   T.	
   A.	
   (1993)	
   CRYSTALLOGRAPHIC	
   STUDIES	
  
ON	
   A	
   FAMILY	
   OF	
   CELLULAR	
   LIPOPHILIC	
   TRANSPORT	
   PROTEINS	
   -­‐	
   REFINEMENT	
   OF	
   P2	
  
MYELIN	
   PROTEIN	
   AND	
   THE	
   STRUCTURE	
   DETERMINATION	
   AND	
   REFINEMENT	
   OF	
  
CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   IN	
   COMPLEX	
   WITH	
   ALL-­‐TRANS-­‐RETINOL,	
   Journal	
  
of	
  Molecular	
  Biology	
  230,	
  1225-­‐1246.	
  
Vasileiou,	
   C.,	
   Vaezeslami,	
   S.,	
   Crist,	
   R.	
   M.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2007)	
   Protein	
   design:	
   reengineering	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   into	
   a	
  
rhodopsin	
  protein	
  mimic,	
  J	
  Am	
  Chem	
  Soc	
  129,	
  6140-­‐6148.	
  
Otwinowski,	
   Z.,	
   and	
   Minor,	
   W.	
   (1997)	
   Processing	
   of	
   X-­‐ray	
   diffraction	
   data	
   collected	
   in	
  
oscillation	
  mode,	
  Macromolecular	
  Crystallography,	
  Pt	
  A	
  276,	
  307-­‐326.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (1997)	
   MOLREP:	
   an	
   automated	
   program	
   for	
   molecular	
  
replacement,	
  Journal	
  of	
  Applied	
  Crystallography	
  30,	
  1022-­‐1025.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (2010)	
   Molecular	
   replacement	
   with	
   MOLREP,	
   Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  22-­‐25.	
  
Murshudov,	
   G.	
   N.,	
   Skubak,	
   P.,	
   Lebedev,	
   A.	
   A.,	
   Pannu,	
   N.	
   S.,	
   Steiner,	
   R.	
   A.,	
   Nicholls,	
   R.	
   A.,	
  
Winn,	
   M.	
   D.,	
   Long,	
   F.,	
   and	
   Vagin,	
   A.	
   A.	
   (2011)	
   REFMAC5	
   for	
   the	
   refinement	
   of	
  
macromolecular	
   crystal	
   structures,	
   Acta	
   Crystallographica	
   Section	
   D-­‐Biological	
  
Crystallography	
  67,	
  355-­‐367.	
  	
  
27	
  

Chapter	
  II:	
  Domain	
  swapping	
  in	
  hCRBPII	
  
	
  

II-­‐1	
  Introduction	
  
Protein	
   oligomers	
   have	
   evolved	
   because	
   of	
   their	
   advantages	
   over	
   their	
   monomers.	
  
These	
   advantages	
   include	
   the	
   possibility	
   of	
   allosteric	
   control,	
   larger	
   binding	
   surfaces,	
   higher	
  
concentration	
   of	
   active	
   sites,	
   new	
   active	
   sites	
   at	
   subunit	
   interfaces	
   and	
   higher	
   stability.	
  
However,	
  the	
  mechanisms	
  for	
  the	
  evolution	
  and	
  assembly	
  of	
  oligomeric	
  species	
  during	
  protein	
  
synthesis	
   remain	
   unclear.	
   Different	
   mechanisms	
   have	
   been	
   proposed	
   for	
   the	
   evolution	
   of	
  
protein	
   oligomers,	
   three-­‐dimensional	
   (3D)	
   domain	
   swapping	
   is	
   among	
   them.	
   (37-­‐42)	
   Three-­‐
dimensional	
   (3D)	
   domain	
   swapping	
   (DS)	
   is	
   a	
   process	
   in	
   which	
   two	
   or	
   more	
   monomer	
   protein	
  
molecules	
   exchange	
   their	
   identical	
   structural	
   elements	
   to	
   form	
   dimers	
   or	
   higher-­‐order	
  
oligomers	
  (Figure	
  II-­‐1).	
  (37-­‐42)	
  	
  
	
  
	
  

HOOC

Closed
Interface

Hinge Loop

Open
Interface

Domain Swapped

	
  

Dimerization
H2N

	
  

NH2

COOH

Monomer
H2N

	
  

COOH

Domain Swapped Dimer

Figure	
   II-­‐1:	
   Cartoon	
   illustration	
   of	
   3D	
   domain	
   swapping.	
   Identical	
  
	
   structural	
  units	
  swap	
  between	
  2	
  monomers,	
  followed	
  by	
  creating	
  the	
  open	
  
interface	
  between	
  them.	
  

28	
  

The	
   region	
   that	
   differs	
   markedly	
   between	
   monomer	
   and	
   oligomer,	
   called	
   the	
   “switch	
  
region”	
  or	
  hinge	
  region”	
  is	
  relatively	
  small.	
  In	
  most	
  cases	
  it	
  consists	
  of	
  a	
  loop	
  in	
  the	
  monomer	
  
that	
  becomes	
  straightened	
  in	
  the	
  DS	
  oligomer.	
  In	
  many	
  cases	
  it	
  is	
  this	
  region	
  that	
  seems	
  to	
  be	
  
the	
  most	
  important	
  determinant	
  for	
  the	
  occurrence	
  of	
  domain	
  swapping.	
  It	
  was	
  first	
  discovered	
  
by	
  Eisenberg	
  and	
  coworkers	
  in	
  1994	
  to	
  account	
  for	
  the	
  dimerization	
  mechanism	
  of	
  diphtheria	
  
toxin	
  (DT)	
  (Figure	
   II-­‐2)	
  and	
  was	
  later	
  shown	
  to	
  be	
  responsible	
  for	
  the	
  dimerization	
  of	
  lyophilized	
  
Bovine	
   pancreatic	
   ribonuclease	
   (Rnase	
   A,	
   which	
   can	
   have	
   either	
   N-­‐terminal	
   or	
   C-­‐terminal	
  
swapped	
  dimers).	
  So	
  far	
  only	
  about	
  40	
  distinctly	
  different	
  proteins	
  have	
  been	
  observed	
  to	
  form	
  
domain	
  swapped	
  dimers.	
  (43-­‐45)	
  
Diphtheria	
   toxin	
   (DT)	
   is	
   a	
   533-­‐residues	
   protein	
   toxin	
   secreted	
   from	
   a	
   bacterium	
   that	
  
causes	
  diphtheria.	
  DT	
  has	
  three	
  domains:	
  a	
  catalytic	
  domain	
  (C),	
  a	
  transmembrane	
  domain	
  (T)	
  
and	
  a	
  receptor-­‐binding	
  domain	
  (R)	
   (37-­‐42).	
  The	
  dimeric	
  form	
  of	
  DT	
  does	
  not	
  form	
  spontaneously	
  
even	
  at	
  higher	
  concentrations.	
  However,	
  DT	
  can	
  be	
  induced	
  to	
  form	
  dimers	
  by	
  decreasing	
  pH,	
  
which	
   converts	
   monomeric	
   DT	
   into	
   open	
   monomers,	
   which	
   form	
   a	
   DT	
   dimer	
   at	
   high	
  
concentration	
  by	
  swapping	
  the	
  globular	
  domain	
  (R	
  domain)	
  (37-­‐42)	
  The	
  structure	
  of	
  monomer	
  DT	
  
revealed	
   the	
   mechanism	
   by	
   which	
   low	
   pH	
   could	
   trigger	
   changes	
   in	
   monomer	
   DT	
   and	
   thereby	
  
form	
   an	
   open	
   monomer.	
   In	
   the	
   monomer,	
   the	
   interdomain	
   interface	
   between	
   the	
   R	
   domain	
  
and	
  C	
  domain	
  is	
  charged,	
  with	
  nine	
  basic	
  and	
  three	
  acidic	
  residues	
  on	
  the	
  R	
  domain	
  interface	
  
and	
  seven	
  acidic	
  residues	
  on	
  the	
  C	
  domain	
  interface.	
  There	
  are	
  three	
  salt	
  bridges	
  stabilizing	
  the	
  
interface	
   at	
   neutral	
   pH.	
   The	
   decrease	
   in	
   pH	
   causes	
   protonation	
   of	
   those	
   acidic	
   residues	
   and	
  
disruption	
   of	
   the	
   salt	
   bridges.	
   Furthermore,	
   buried	
   positive	
   charges	
   in	
   the	
   interface	
   favor	
   the	
  
open	
   monomer	
   formation.	
   During	
   the	
   dimerization	
   of	
   DT,	
   higher	
   order	
   oligomers	
   were	
  
29	
  

observed	
  by	
  size	
  exclusion	
  HPLC.	
  These	
  oligomers	
  include	
  trimers,	
  tetramers	
  and	
  pentamers	
  (37-­‐
42)

.	
  In	
  most	
  cases	
  the	
  exchanged	
  region	
  is	
  located	
  at	
  the	
  N-­‐	
  or	
  C-­‐terminus	
  of	
  the	
  protein,	
  though	
  

the	
  exchanged	
  part	
  is	
  found	
  in	
  the	
  middle	
  of	
  the	
  sequence	
  in	
  a	
  few	
  instances.	
  (46-­‐49)	
  
	
  
a	
  
	
  
	
  

In	
  acidic	
  
pH	
  

	
  
	
  

Monomer	
  	
  DT	
  

Dimer	
  	
  DT	
  

	
  
	
  

b	
  

	
  
	
  
	
  
Figure	
   II-­‐2:	
   a.	
   Left:	
   DT	
   monomer	
   (PDB	
   ID	
   4OW6).	
   Right:	
   DT	
   dimer	
   (4AE1).	
   The	
   two	
  
	
  
subunits	
  
are	
  blue	
   and	
  pink.	
  In	
  acidic	
  pH	
  equilibrium	
   goes	
  toward	
  DS	
  dimer	
   formation.	
  
b.	
  Left	
  structure	
  of	
  DS	
  dimer	
  of	
  RnaseA	
  (PDB	
  1A2W)	
  and	
  cyclic	
  DS	
  trimes	
  (PDB	
  5SRA).	
  
Domain	
   swapping,	
   combined	
   with	
   gene	
   duplication,	
   has	
   been	
   postulated	
   to	
   be	
  
responsible	
  for	
  the	
  evolution	
  of	
  a	
  number	
  of	
  larger	
  protein	
  domains	
  from	
  smaller	
  fragments(50).	
  
Domain	
   swapping	
   can	
   also	
   lead	
   to	
   aggregation,	
   amyloid	
   and	
   fibril	
   formation,	
   which	
   lead	
   to	
  

30	
  

protein	
   malfunction(51).	
   Protein	
   misfolding	
   and	
   aggregation	
   is	
   the	
   cause	
   of	
   a	
   number	
   of	
  
maladies	
  including	
  Parkinson’s	
  disease,	
  Alzheimer	
  disease,	
  diabetes,	
  Huntington’s	
  disease	
  and	
  
many	
   others(52-­‐58).	
   	
   Triggers	
   of	
   domain-­‐swapping	
   include	
   lyophilization,	
   non-­‐physiological	
   pH,	
  
unusually	
   high	
   protein	
   concentration	
   (59,	
   60)	
   (43,	
   61)	
   temperature,	
   mutation	
   and	
   even	
   ligand	
  
binding(62,	
  63).	
  	
  
The	
  classic	
  example	
  of	
  domain	
  swapping	
  is	
  bovine	
  pancreatic	
  ribonuclease	
  A	
  (RNase	
  A),	
  which,	
  
the	
   DS	
   of	
   the	
   N-­‐terminal	
   exchange,	
   was	
   proposed	
   in	
   1962	
   by	
   Crestfield,	
   Stein,	
   and	
   Moore	
   to	
  
describe	
  its	
  behavior	
  under	
  acidic	
  conditions.	
  Followed	
  by	
  the	
  first	
  X-­‐ray	
  structures	
  for	
  a	
  domain	
  
swapped	
  dimer	
  in	
  the	
  late	
  nineties.	
  In	
  RNase	
  case,	
  either	
  the	
  N-­‐terminal	
  helix	
  or	
  the	
  C-­‐terminal	
  
strand	
  can	
  domain	
  swap.	
  The	
  swapped	
  unit	
  and	
  its	
  oligomeric	
  state	
  can	
  vary	
  significantly	
  and	
  
RNase	
   A	
   illustrates	
   the	
   possibilities	
   of	
   domain	
   swapping	
   in	
   the	
   trimer	
   both	
   N-­‐	
   and	
   C-­‐terminal	
  
units	
   are	
   swapped	
   and	
   a	
   circular	
   arrangement	
   succeeds	
   (Figure	
   II-­‐2b).	
   Another	
   example	
   is	
  
Cyanovirin-­‐N	
  (CV-­‐N),	
  which	
  is	
  a	
  101	
  amino	
  acid	
  cyanobacterial	
  lectin.	
  CV-­‐N	
  inactivates	
  HIV	
  and	
  
is	
   a	
   general	
   virucidal	
   agent	
   against	
   other	
   enveloped	
   viruses.	
   The	
   original	
   solution	
   structure	
   was	
  
a	
   monomer,	
   whereas	
   the	
   subsequently	
   determined	
   X-­‐ray	
   structures	
   were	
   domain	
   swapped	
  
dimers,	
  which	
  each	
  domain	
  comprises	
  a	
  triple-­‐stranded	
  β-­‐sheet	
  with	
  a	
  β-­‐hairpin	
  packed	
  on	
  top.	
  
A	
  helical	
  linker	
  is	
  located	
  in	
  the	
  middle	
  of	
  the	
  sequence.	
  It	
  was	
  shown	
  that	
  the	
  DS	
  dimer	
  is	
  a	
  
kinetically	
   trapped	
   folding	
   intermediate	
   at	
   high	
   protein	
   concentrations	
   that	
   convert	
   into	
   the	
  
slightly	
   more	
   stable	
   monomer	
   form	
   at	
   physiological	
   (>30oC)	
   temperature.1d,	
   10	
   At	
   room	
  
temperature	
   or	
   below,	
   however,	
   the	
   dimer	
   lifetime	
   is	
   sufficiently	
   long	
   for	
   structural	
  
characterization	
  in	
  solution.	
  The	
  fact	
  that	
  both	
  monomeric	
  and	
  domain-­‐swapped	
  dimeric	
  CV-­‐N	
  
coexist	
  in	
  solution	
  under	
  identical	
  conditions	
  indicates	
  that	
  the	
  free	
  energies	
  of	
  folding	
  for	
  both	
  
31	
  

quaternary	
   states	
   must	
   be	
   comparable	
   and	
   the	
   kinetic	
   barrier	
   between	
   the	
   monomer	
   and	
  
dimer	
   has	
   to	
   be	
   significant.	
   This	
   can	
   be	
   altered	
   by	
   mutation	
   of	
   residues	
   in	
   the	
   hinge	
   region.	
   For	
  
instance,	
   changing	
   the	
   proline	
   in	
   the	
   hinge	
   region	
   to	
   glycine	
   resulted	
   in	
   a	
   substantial	
  
stabilization	
   of	
   this	
   monomeric	
   P51G	
   mutant	
   by	
   >5	
   kcal/mol	
   compared	
   to	
   wild-­‐type.	
   A	
   S52P	
  
mutant	
   yielded	
   predominantly	
   dimeric	
   protein	
   due	
   to	
   destabilization	
   of	
   the	
   monomer.	
   More	
  
and	
  more	
  domain	
  swapped	
  protein	
  structures	
  are	
  becoming	
  elucidated,	
  and,	
  for	
  several	
  cases,	
  
growing	
   evidence	
   supports	
   that	
   the	
   dimer	
   or	
   multimer	
   is	
   an	
   active,	
   biological	
   important	
  
structure.	
   Regardless	
   of	
   whether	
   domain	
   swapping	
   is	
   a	
   specific	
   mechanism	
   for	
   regulation	
   in	
  
vivo,	
   it	
   is	
   becoming	
   clear	
   that	
   domain	
   swapping	
   is	
   a	
   means	
   by	
   which	
   stable	
   oligomers	
   can	
   be	
  
generated	
  under	
  evolutionary	
  force.	
  By	
  protein	
  expression	
  or	
  refolding	
  at	
  high	
  concentrations,	
  
suggesting	
   that	
   high	
   concentrations	
   of	
   folding	
   intermediates	
   may	
   be	
   involved	
   in	
   the	
   process.(64,	
  
65)

	
   	
   So	
   far	
   no	
   case	
   has	
   been	
   observed	
   where	
   both	
   the	
   monomer	
   and	
   dimer	
   of	
   a	
   wild-­‐type	
  

protein	
   are	
   expressed	
   without	
   rapid,	
   equilibrating	
   monomer/dimer	
   exchange.	
   Domain-­‐
Swapping	
   can	
   also	
   be	
   caused	
   by	
   mutation.	
   For	
   example	
   a	
   single	
   site	
   mutation	
   in	
   protein	
   L	
   from	
  
Peptostreptococcus	
  Magnus	
  leads	
  to	
  3D	
  domain	
  swapping	
  (66).	
  	
  
As	
   discussed	
   in	
   the	
   previous	
   chapter,	
   iLBPs	
   are	
   relatively	
   small	
   (126-­‐140	
   amino	
   acid,	
  
predominantly	
  β	
  sheet	
  proteins)	
  cytosolic	
  proteins.	
  Structures	
  of	
  a	
  number	
  of	
  iLBPs	
  have	
  been	
  
determined	
   and	
   all	
   share	
   this	
   monomeric	
   fold.	
   	
   (Figure	
   II-­‐1)	
   Numerous	
   studies	
   of	
   the	
   folding	
  
mechanism	
  of	
  members	
  of	
  the	
  iLBP	
  family	
  have	
  been	
  conducted,	
  specially	
  focused	
  on	
  human	
  
Cellular	
   Retinoic	
   Acid	
   Binding	
   Protein	
   I	
   (hCRABPI).	
   (67-71)	
   	
   In	
   an	
   elegant	
   series	
   of	
   experiments,	
  
using	
   a	
   variety	
   of	
   techniques,	
   the	
   Gierasch	
   lab	
   has	
   concluded	
   that	
   CRABPI	
   undergoes	
   early	
   beta	
  
barrel	
  collapse.	
  

32	
  

They	
   used	
   CRAPI	
   with	
   its	
   simple	
   architecture	
   to	
   study	
   structural	
   basis	
   of	
   β	
   sheet	
   proteins	
  
formation.	
   Regarding	
   this	
   purpose,	
   three	
   only	
   Tryptophan	
   residues,	
   with	
   structurally	
   distinct	
  
locations	
   as	
   three	
   probes,	
   were	
   mutated.	
   Folding	
   of	
   these	
   mutants	
   was	
   examined	
   using	
  
stopped-­‐flow	
   fluorescence	
   and	
   circular	
   dichroism.	
   As	
   a	
   result	
   within	
   10	
   ms,	
   the	
   rapid	
  
hydrophobic	
   collapse	
   occurs	
   and	
   stopped-­‐flow	
   circular	
   dichroism	
   shows	
   significant	
   secondary	
  
structure	
   content	
   and	
   it	
   (chain	
   topology)	
   gets	
   developed	
   by	
   100	
   ms.	
   Then	
   followed	
   by	
   the	
  
specific	
   packing	
   of	
   the	
   β-­‐sheet	
   sidechains	
   and	
   	
   formation	
   of	
   the	
   native	
   hydrogen-­‐bond	
   network	
  
(T	
  =	
   1	
   s).	
   These	
   are	
   hallmarks	
   of	
   the	
   pathway,	
   thus	
   preventing	
   off	
   pathway	
   aggregation	
   and	
  
potential	
  amyloid	
  formation	
  in	
  this	
  protein.(72)	
  
Investigations	
   into	
   the	
   folding	
   mechanism	
   of	
   other	
   members	
   of	
   the	
   family	
   have	
   also	
  
been	
   conducted,	
   all	
   of	
   which	
   are	
   consistent	
   with	
   a	
   relatively	
   rugged	
   energy	
   landscape,	
  
indicating	
  a	
  number	
  of	
  possible	
  stable	
  or	
  meta-­‐stable	
  intermediates,	
  a	
  situation	
  that	
  is	
  generally	
  
characteristic	
  of	
  proteins	
  that	
  are	
  largely	
  beta	
  sheet	
  in	
  structure.(72)	
  	
  
Also	
   in	
   chapter	
   II	
   was	
   mentioned	
   that	
   CRBPs	
   belong	
   to	
   the	
   iLBP	
   family.	
   (6,	
   10-­‐13)	
   The	
  
combination	
  of	
  small	
  size	
  and	
  relatively	
  large	
  binding	
  cavity	
  made	
  these	
  family	
  members	
  ideal	
  
templates	
  for	
  use	
  in	
  protein	
  design	
  applications.	
  
The	
   Geiger	
   and	
   Borhan	
   labs	
   have	
   used	
   these	
   templates	
   for	
   a	
   wide-­‐ranging	
   protein	
   design	
  
applications	
   for	
   more	
   than	
   a	
   decade.	
   First	
   both	
   hCRABPII	
   and	
   hCRBPII	
   were	
   modified	
   to	
   be	
  
Rhodopsin	
   mimics	
   that	
   bind	
   and	
   react	
   with	
   retinal	
   to	
   form	
   the	
   protonated	
   Schiff	
   base	
   (PSB)	
  
which	
  is	
  found	
  in	
  the	
  binding	
  cavity	
  of	
  all	
  rhodopsins.(31,	
  73)	
  Then	
  the	
  environment	
  around	
  the	
  
bound	
  chromophore	
  was	
  modified	
  to	
  study	
  the	
  mechanisms	
  of	
  absorption	
  wavelength	
  tuning	
  
that	
   give	
   rise	
   to	
   color	
   vision,	
   and	
   the	
   light	
   selectivity	
   of	
   all	
   the	
   other	
   rhodopsins,	
   which	
   is	
  

33	
  

required	
  for	
  their	
  function.(74-­‐76)	
  Rhodopsin	
  mimics	
  that	
  photo-­‐isomerize	
  both	
  in	
  solution	
  and	
  in	
  
the	
  crystal	
  were	
  created	
  to	
  allow	
  the	
  detailed	
  study	
  of	
  isomerization	
  processes	
  in	
  rhodopsins.(77,	
  
78)

	
   In	
   our	
   collaboration,	
   we	
   have	
   also	
   created	
   a	
   new	
   class	
   of	
   fluorescent	
   proteins	
   by	
   binding	
  

fluorophores	
   in	
   the	
   binding	
   pocket	
   that	
   also	
   react	
   to	
   form	
   PSB’s	
   and	
   allow	
   regulation	
   of	
   the	
  
emission	
  wavelength.	
   (79)	
  (Some	
  of	
  this	
  work	
  is	
  described	
  further	
  in	
  chapter	
  IV)	
  To	
  perform	
  all	
  
the	
  above-­‐mentioned	
  projects,	
  we	
  had	
  to	
  make,	
  express	
  and	
  evaluate	
  hundreds	
  of	
  mutants	
  of	
  
these	
  proteins.	
  During	
  the	
  preparation	
  of	
  mutants	
  for	
  these	
  studies,	
  Dr.	
  Wang	
  and	
  Dr.	
  Nossoni	
  
for	
   the	
   fist	
   time	
   observed	
   protein	
   bands	
   that	
   eluted	
   separately	
   on	
   ion-­‐exchange	
  
chromatography,	
   but	
   nonetheless,	
   were	
   the	
   same	
   species	
   on	
   SDS-­‐PAGE.	
   	
   This	
   led	
   us	
   to	
   the	
  
discovery	
  of	
  a	
  set	
  of	
  hCRBPII	
  protein	
  mutants	
  capable	
  of	
  domain	
  swapped	
  dimerization.	
  It	
  was	
  a	
  
surprise	
   to	
   discover	
   that	
   some	
   hCRBPII	
   mutants	
   predominantly	
   form	
   dimers	
   via	
   domain	
  
swapping	
  instead	
  of	
  the	
  monomeric	
  form.	
  (45)	
  (Figure	
  II-­‐3)	
  
	
  
	
  
	
  

	
  

2 equiv

	
  	
  

	
  
	
  

nomer" 	
  
rmediate 	
  

domain swapped dimer

Figure	
  II-­‐3:	
  The	
  hCRBPII	
  DS	
  dimer.	
  

We	
   were	
   the	
   first	
   group	
   to	
   report	
   domain	
   swapped	
   structures	
   of	
   hCRBPII. (45)	
   The	
   extent	
  
of	
   domain	
   swapping	
   is	
   large	
   (almost	
   50%	
   of	
   the	
   protein	
   sequence)	
   and	
   occurs	
   during	
   routine	
  

34	
  

protein	
  bacterial	
  over-­‐expression	
  at	
  16-­‐30	
  °C.	
  HCRBPII	
  domain	
  swapped	
  dimerization	
  is	
  favored	
  
by	
  high	
  expression	
  levels	
  in	
  E.	
  coli	
  and	
  high	
  concentrations	
  in	
  vitro.	
  	
  Most	
  importantly	
  protein	
  
stability	
  studies	
  show	
  the	
  domain	
  swapped	
  dimer	
  to	
  be	
  a	
  less	
  stable,	
  kinetically	
  trapped	
  species	
  
(depending	
  on	
  the	
  mutant).	
  The	
  resultant	
  dimer	
  structure	
  represents	
  a	
  fold	
  not	
  identified	
  in	
  the	
  
SCOP	
  or	
  CATH	
  protein	
  fold	
  databases.	
  	
  Together,	
  our	
  data	
  supports	
  the	
  hypothesis	
  that	
  folding	
  of	
  
hCRBPII	
  occurs	
  via	
  a	
  structurally	
  ordered	
  “open	
  monomer,”	
  with	
  the	
  relative	
  orientation	
  of	
  the	
  
N-­‐	
   and	
   C-­‐	
   terminal	
   structural	
   units	
   determining	
   the	
   propensity	
   for	
   dimerization.2c	
   The	
   existence	
  
of	
   this	
   dimer	
   for	
   hCRBPII	
   suggests	
   a	
   folding	
   pathway	
   distinct	
   from	
   that	
   characterized	
   for	
  
hCRABPI,	
  indicating	
  that	
  evolutionary	
  related	
  members	
  of	
  the	
  same	
  structural	
  family	
  may	
  follow	
  
different	
  folding	
  pathways.	
  	
  
Later	
   on,	
   we	
   determined	
   structures	
   of	
   holo	
   hCRBPII	
   domain	
   swapped	
   dimer	
   variants.	
  
Together,	
   these	
   structures	
   show	
   an	
   extremely	
   large	
   and	
   reproducible	
   change	
   in	
   the	
   relative	
  
orientation	
   of	
   the	
   two	
   domains	
   of	
   the	
   dimer	
   upon	
   ligand	
   binding,	
   inviting	
   the	
   possibility	
   that	
  
iLBP	
  domain	
  swapped	
  dimers	
  could	
  be	
  allosterically-­‐regulated	
  forms	
  of	
  these	
  proteins,	
  at	
  least	
  
in	
   some	
   cases.	
   Ideally,	
   the	
   domain	
   swapped	
   dimer	
   may	
   represent	
   a	
   new,	
   biologically	
   relevant	
  
structural	
   fold	
   for	
   the	
   iLBP	
   protein	
   family,	
   though	
   it	
   remains	
   to	
   be	
   seen	
   if	
   they	
   form	
   in	
   vivo,	
   and	
  
have	
  distinct	
  functionality	
  relative	
  to	
  their	
  parent,	
  monomeric	
  forms.	
  	
  

II-­‐2	
  Origin	
  of	
  dimerization	
  in	
  hCRBPII	
  
	
  

As	
   it	
   was	
   mentioned	
   above,	
   we	
   developed	
   hCRBPII	
   as	
   rhodopsin	
   mimics	
   to	
   study	
   the	
  

mechanism	
  of	
  wavelength	
  tuning.	
  To	
  this	
  purpose,	
  the	
  binding	
  site	
  was	
  redesigned	
  to	
  bind	
  all-­‐
trans	
  retinal	
  and	
  form	
  a	
  PSB	
  via	
  the	
  Q108K:K40L	
  (KL)	
  double	
  mutant	
  by	
  Dr.	
  Wang	
  previously	
  (80).	
  	
  	
  

35	
  

	
  
	
  

2.53 Å

	
  
	
  
4.23 Å

	
  
	
  
	
  
	
  

Figure	
  II-­‐4:	
  Trajectory	
  of	
  Tyr	
  60,	
   Glu	
  72	
  (both	
  shown	
  as	
  stick	
  representations,	
  
colored	
  blue)	
  in	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinal.	
  

The	
  residues	
  situated	
  around	
  the	
  chromophore	
  were	
  then	
  selected	
  for	
  systematic	
  mutation	
  in	
  
an	
  effort	
  to	
  understand	
  the	
  mechanism	
  of	
  wavelength	
  tuning.	
  Tyr60,	
  conserved	
  only	
  in	
  CRBPs,	
  
is	
  located	
  in	
  beta-­‐strand	
  C	
  and	
  approximately	
  4Å	
  from	
  the	
  center	
  of	
  the	
  polyene	
  chain	
  of	
  all-­‐
trans-­‐retinal.	
  (Figure	
  II-­‐4)	
  (17, 80-82).	
  	
  
The	
   mutant	
   Q108K:K40L:Y60W	
   (KLY60W)	
   was	
   generated	
   by	
   Dr.	
   Wang,	
   expressed	
   and	
  
purified	
   as	
   previously	
   described	
   in	
   the	
   hopes	
   of	
   producing	
   a	
   π-­‐π	
   interaction	
   with	
   the	
  
chromophore	
   that	
   could	
   potentially	
   delocalize	
   the	
   positive	
   charge	
   of	
   the	
   PSB	
   along	
   the	
  
chromophore,	
  leading	
  to	
  shifted	
  spectra.(80)	
  In	
  its	
  purification	
  process,	
  Source-­‐Q	
  ion	
  exchange	
  
chromatography	
   (IEX)	
   unexpectedly	
   gave	
   two	
   KLY60W	
   hCRBPII-­‐containing	
   protein	
   peak	
  
fractions,	
   one	
   eluting	
   at	
   80mM	
   NaCl,	
   where	
   hCRBPII	
   mutants	
   typically	
   elute,	
   and	
   a	
   second	
  
eluting	
  at	
  160	
  mM	
  NaCl	
   (45)	
  (Figure	
   II-­‐5a).	
  The	
  UV-­‐Vis	
  spectra	
  of	
  the	
  all-­‐trans-­‐retinylidene	
  PSB	
  
complex	
  with	
  protein	
  from	
  each	
  of	
  the	
  fractions	
  gave	
  distinct	
  spectra.	
  The	
  80	
  mM	
  and	
  160	
  mM	
  
protein	
   fractions,	
   when	
   bound	
   to	
   retinal,	
   showed	
   Νmax	
   of	
   496nm	
   and	
   514nm,	
   respectively	
  
(Figure	
   II-­‐5b,	
   5c).	
   	
   At	
   lower	
   expression	
   temperature	
   (25°C)	
   significantly	
   more	
   of	
   the	
   80	
   mM	
  

36	
  

fraction	
   was	
   produced	
   while	
   a	
   higher	
   expression	
   temperature	
   (30°C)	
   gave	
   similar	
   amounts	
   of	
  
each	
  species	
  (Table	
  II-­‐1)	
  	
  
	
  

a	
  
	
  

b	
  

c	
  
	
  

d	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  

e	
  

f	
  

	
  
	
  
	
  
	
  

g	
  
	
  

h	
  
	
  

i	
  

	
  
	
  
	
  

Figure	
  II-­‐5:	
  a.	
  IEX	
  chromatogram	
  of	
  the	
  KLY60W,	
  monitored	
  at	
  280nm.	
  b.	
  The	
  UV-­‐
Vis	
   spectra	
   of	
   the	
   80mM	
   salt	
   elution	
   (red)	
   and	
   the	
   160mM	
   salt	
   elution	
   (blue)	
   of	
  
KLY60W	
   incubated	
   with	
   all-­‐trans-­‐retinal.	
   c.	
   The	
   UV-­‐Vis	
   spectra	
   of	
   the	
   refolded	
  
80mM	
   eluent	
   (low	
   salt,	
   blue)	
   after	
   denaturation	
   and	
   refolding	
   at	
   0.03	
   mM	
  
concentration	
   and	
   the	
   native	
   160mM	
   (high	
   salt,	
   green)	
   and	
   native	
   80mM	
   elution	
  
(red),	
  all	
  bound	
  to	
  all-­‐trans-­‐retinal.	
  d.	
  The	
  size	
  exclusion	
  chromatogram	
  (Superdex	
  
200	
  16/75	
  column)	
  for	
  dimer	
  (the	
  160mM	
  elution)	
  KLY60W,	
  after	
  maintaining	
  the	
  
protein	
   at	
   room	
   temperature	
   for	
   several	
   days,	
   showing	
   only	
   the	
   monomer	
   size	
  
peak.	
  e.	
  Gel	
  filtration	
  chromatogram	
  of	
  monomer	
  (the	
  80mM	
  elution)	
  of	
  KLY60W,	
  
showing	
   only	
   the	
   dimer	
   fractions.	
   f-­‐i.	
   Ion	
   exchange	
   chromatographs	
   of	
   WT	
   and	
  
various	
  hCRBPII	
  mutants,	
  monitored	
  at	
  280	
  nm.	
  f.	
  Y60I	
  at	
  30°C.	
  g.	
  Y60L	
  at	
  30°C.	
  h.	
  
Y60L	
  at	
  25°C.	
  i.	
  WT	
  hCRBPII	
  expressed	
  at	
  30	
  °C.	
  

37	
  

Size-­‐exclusion	
  chromatography	
  (SEC)	
  showed	
  that	
  the	
  80	
  mM	
  protein	
  eluted	
  as	
  a	
  15kD	
  
monomer,	
   and	
   the	
   160	
   mM	
   fraction	
   eluted	
   as	
   a	
   30kD	
   dimer.	
   No	
   equilibrium	
   between	
   the	
  
monomer	
   and	
   dimer	
   forms	
   was	
   observed,	
   even	
   after	
   incubation	
   for	
   10	
   days	
   at	
   room	
  
temperature	
  (Figure	
   II-­‐5d	
   and	
   5e).	
  Each	
  species	
  has	
  been	
  tested	
  to	
  achieve	
  more	
  information	
  
about	
   their	
   stability.	
   Their	
   melting	
   curves	
   showed	
   the	
   dimer	
   to	
   be	
   less	
   stable	
   than	
   the	
  
monomer	
   (denaturing	
   at	
   69°C,	
   explained	
   in	
   II-­‐7-­‐4	
   section,	
   Figure	
   II-­‐6c	
   and	
   31),	
   however	
   the	
  
monomer	
  was	
  stable	
  to	
  95°C	
  (Figure	
  II-­‐6f	
  and	
  II-­‐31).	
  The	
  160	
  mM	
  salt	
  fraction	
  was	
  denatured	
  
with	
  8M	
  urea	
  and	
  refolded	
  at	
  low	
  concentration	
  (0.03	
  mM)	
  (Table	
   II-­‐1).	
  The	
  UV-­‐Vis	
  spectrum	
  of	
  
the	
   retinal	
   PSB-­‐bound	
   refolded	
   monomer	
   was	
   identical	
   to	
   the	
   originally	
   isolated	
   monomer,	
  
indicating	
  the	
  refolded	
  proteins	
  to	
  be	
  structurally	
  similar	
  (Figure	
  II-­‐5c).	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

38	
  

	
  

0

0

-20

-10

-30

-20

-40
-60

Wild-Type Dimer
Tm = 52 ÂşC

-80

-40
Y60L-Dimer
Tm = 56 ÂşC

-50

	
  

KLY60W-Dimer
Tm = 69 ÂşC

-50

-60
-70

-70

-140
20 30 40 50 60 70 80 90 100

-80
20 30 40 50 60 70 80 90 100

-80
20

-60

Temperature (ÂşC)

d	
  

e	
   -25

-50

	
  

-65
-70
-75
-80

-20

-35
-40
Y60L-Monomer

-45

-55

20

40

60

80

100

Temperature (ÂşC)

120

-10

120

-30
-40

KLY60W- Monomer

	
  

-50
-60

-50

-85
-90

60
80
100
Temperature (ÂşC)

-30
CD (mdeg)

Wild-Type Monomer

40

f	
  
CD (mdeg)

-55
CD (mdeg)

-40

-120

-60

	
  

-30

-100

Temperature (ÂşC)

	
  

c	
  

-10

-20

	
  

C

20

CD (mdeg)

	
  

b	
  

B

CD (mdeg)

	
  

a	
  A

CD (mdeg)

	
  

-70

20

40

60

80

100

120

30 40 50 60 70 80 90 100 110
Temperature (ÂşC)

Temperature (ÂşC)

Figure	
  II-­‐6:	
  Thermal	
  melting	
  curves:	
  a.	
  WT	
  hCRBPII	
  dimer	
  (θ	
  measured	
  at	
  220	
  nm),	
  Tm	
  =	
  
52	
   °C;	
  b.	
  Y60L	
  dimer	
   (θ	
  measured	
   at	
  227	
   nm),	
  Tm	
   =	
  56	
  °C;	
  c.	
  KLY60W	
   dimer	
  (θ	
  measured	
  
at	
   220	
   nm),	
   Tm	
   =	
   69	
   °C.	
   d.	
   WT	
   hCRBPII	
   monomer	
   (θ	
   measured	
   at	
   220	
   nm);	
   e.	
   Y60L	
  
monomer	
  (θ	
  measured	
  at	
  215	
  nm);	
  f.	
  KLY60W	
  monomer	
  (θ	
  measured	
  at	
  220	
  nm).	
  Note	
  
that	
   most	
   of	
   the	
   secondary	
   structure	
   is	
   preserved	
   in	
   both	
   the	
   WT	
   and	
   KLY60W	
  
monomers,	
   even	
   at	
   the	
   highest	
   temperature	
   achieved	
   (CD	
   intensities	
   of-­‐65mdeg	
   and	
   -­‐
50mdeg	
   respectively).	
   Thus	
   melting	
   curves	
   can	
   only	
   give	
   a	
   lower	
   limit	
   for	
   the	
   TM,	
  
consistent	
   with	
   the	
   view	
   that	
   the	
   monomers	
   are	
   more	
   stable	
   than	
   the	
   dimers.	
   The	
  
apparent	
  lack	
  of	
  a	
  Tm	
  for	
  the	
  monomers	
  might	
  suggest	
  the	
  presence	
  of	
  thermally	
  induced	
  
folding	
   intermediates.	
   Clearly	
   the	
   unfolding	
   process	
   is	
   not	
   monolithic	
   as	
   that	
   observed	
  
for	
   the	
   dimeric	
   species.	
   This	
   lends	
   further	
   support	
   to	
   the	
   suggestion	
   that	
   the	
   open	
  
monomer,	
   which	
   is	
   presumably	
   obtained	
   from	
   melting	
   of	
   the	
   dimer,	
   has	
   a	
   different	
  
folding	
  trajectory	
  (and	
   consequently	
   a	
   different	
   unfolding	
   trajectory)	
   as	
  compared	
  to	
   the	
  
monomeric	
  species.	
  	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  

II-­‐3	
  Dimer	
  formation	
  of	
  other	
  hCRBPII	
  variants	
  
At	
   first,	
   we	
   hypothesized	
   mutation	
   of	
   the	
   residue	
   60	
   in	
   the	
   KLY60W	
   series	
   was	
   most	
  
likely	
  responsible	
  for	
  the	
  formation	
  of	
  the	
  dimeric	
  species	
  since	
  many	
  hCRBPII	
  proteins	
  mutated	
  
at	
  the	
  40	
  and	
  108	
  positions	
  have	
  been	
  purified	
  in	
  monomeric	
  form	
  in	
  our	
  lab	
  previously(80).	
  To	
  
this	
  purpose,	
  we	
  made	
  several	
  single	
  mutants	
  at	
  this	
  position	
  (Table	
  II-­‐2).	
  	
  

39	
  

Y60W	
  mutant	
  yielded	
  the	
  same	
  monomer/dimer	
  ratios	
  as	
  KLY60W,	
  confirming	
  the	
  importance	
  
of	
   position	
   60.	
   (45)	
   	
   In	
   fact	
   most	
   mutations	
   resulted	
   in	
   increased	
   dimer	
   formation,	
   including	
  
Y60F,	
  Y60L	
  and	
  Y60I,	
  however	
  the	
  Y60H	
  mutant	
  produced	
  no	
  dimeric	
  protein.	
  (Table	
   II-­‐2	
   and	
  
Figure	
   II-­‐5f-­‐h)	
   Together,	
   this	
   data	
   emphasizez	
   the	
   fact	
   that	
   position	
   60	
   mutations	
   can	
  
significantly	
   favor	
   the	
   formation	
   of	
   dimeric	
   species.	
   	
   Position	
   60	
   lies	
   near	
   the	
   N-­‐terminal	
   end	
   of	
  
strand	
   4,	
   and	
   is	
   not	
   a	
   position	
   conserved	
   in	
   iLBP	
   family	
   members,	
   though	
   it	
   is	
   a	
   Tyr	
   in	
   most	
  
retinol	
  binding	
  proteins.(81)	
  	
  
	
  
Table	
  II-­‐
1:	
   The	
   Effect	
   of	
   expression	
   temperature	
   and	
  in	
   vitro	
   refolding	
   protein	
  concentration	
  
	
  
on	
  the	
  dimer/monomer	
  	
  (D/M)	
  ratio.	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  

	
  	
  Protein	
  

D/M	
  (%)a	
   D/M	
  (%)a	
   In	
  vitro	
  
at	
  25	
  °C	
  
at	
  30	
  °C	
   refolding
b

WT-­‐hCRBPII	
  

0/100	
  

10/90	
  

	
  Q108K:K40L:	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Y60W	
  

30/70	
  

60/40	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  Y60L	
  

80/20	
  

90/10	
  

	
  

0.03	
  mM	
  
0.13	
  mM	
  
0.3	
  mM	
  
0.03	
  mM	
  
0.6	
  mM	
  
0.03	
  mM	
  
0.13	
  mM	
  
0.4	
  mM	
  

Percent	
  of	
  
	
  D/M	
  (%)c	
  
0/100	
  
0/100	
  
40/60	
  
0/100	
  
30/70	
  
20/80	
  
40/60	
  
70/30	
  

a.	
  Mass	
  ratio	
  of	
  dimer	
  to	
  monomer;	
  b.	
  Starting	
  concentration;	
  c.	
  Mass	
  ratio,	
  after	
  in	
  vitro	
  refolding.	
  
	
  
	
  

We	
   decided	
   to	
   perform	
   several	
   in	
   vitro	
   refolding	
   experiments	
   on	
   the	
   WT	
   and	
   several	
   of	
   the	
  
mutants	
  (Table	
  II-­‐1	
  and	
  Figure	
  II-­‐7),	
  to	
  better	
  understand	
  the	
  nature	
  of	
  the	
  dimerization	
  in	
  this	
  
protein.	
  	
  
	
  

40	
  

	
  

a	
  

b	
  

c	
  

	
  
	
  
	
  
	
  

d	
  

e	
  

f	
  

	
  
	
  
	
  
Figure	
   II-­‐7:	
   Size	
   exclusion	
   chromatography	
   after	
   protein	
   denaturation	
   and	
   refolding,	
   all	
  
monitored	
  at	
  280	
  nm.	
  a.	
  Y60L	
  mutant	
  refolded	
  at	
  0.03	
  mM	
  concentration.	
   b.	
  Y60L	
  refolded	
  at	
  
0.13	
  mM	
  concentration.	
  c.	
  Y60L	
  refolded	
  at	
  0.4	
  mM	
  concentration.	
  d.	
  	
  WT	
  hCRBPII	
  refolded	
  at	
  
0.03	
   mM	
   concentration.	
   e.	
   WT	
   hCRBPII	
   refolded	
   at	
   0.13	
   mM	
   concentration.	
   f.	
   WT	
   hCRBPII	
  
refolded	
  at	
  0.3	
  mM	
  concentration.	
  	
  
In	
   each	
   case,	
   dimer	
   formation	
   was	
   favored	
   by	
   higher	
   refolding	
   concentrations	
   while	
  
monomer	
  formation	
  was	
  favored	
  at	
  lower	
  refolding	
  concentrations.	
  	
  This	
  correlated	
  with	
  our	
  E.	
  
coli-­‐expression	
  results,	
  where	
  expression	
  at	
  higher	
  temperatures,	
  which	
  leads	
  to	
  higher	
  levels	
  
of	
  expression	
  and	
  therefore	
  higher	
  concentrations	
  of	
  pre-­‐folded	
  proteins,	
  (30	
  °C	
  versus	
  25	
  °C,	
  
Table	
   II-­‐1	
   and	
   Figure	
   II-­‐7).	
  Melting	
  experiments	
  also	
  showed	
  the	
  dimer	
  to	
  be	
  less	
  stable	
  than	
  
the	
  monomer	
  for	
  two	
  of	
  the	
  three	
  variants	
  (WT,	
  KLY60W,	
  Figure	
   II-­‐6a,	
   6c,	
   6d	
   and	
   6f),	
  indicating	
  
the	
  dimer	
  to	
  be	
  the	
  kinetic	
  product	
  (the	
  melting	
  curve	
  for	
   the	
  Y60L	
  monomer	
  was	
  not	
  readily	
  
interpretable,	
  Figure	
  II-­‐6b	
  and	
  6e).	
  (45)	
  
	
  
	
  
	
  

41	
  

!

Table	
  II-­‐2:	
  	
  Monomer/dimer	
  ratio	
  of	
  various	
  residue	
  60	
  mutants.	
  During	
  bacterial	
  expression	
  at	
  
30	
  °C	
  
	
  
Mutations
Dimer/Monomer

mass ratio (%)

	
  

WT

10/90 (no dimer
at RT)

Y60L

90/10

S55W

20/80

S55W:Y60L

60/40

	
  

Y60I

80/20

	
  

Y60Q

20/80

	
  

Y60D

30/70

	
  

Y60T

No dimer

	
  

Y60W

50/50

	
  

Q108K:K40L:Y60F

40/60

	
  

Q108K:K40L:Y60W

60/40

	
  

Δ56:WT

20/80

	
  

F57G:Y60L

70/30

	
  

T56P-Y60L

70/30

	
  

Δ56:Y60L

90/10

	
  

E72A

50/50

	
  

Q108K:K40L:Y60H

No dimer

	
  
	
  
	
  
	
  

	
  
	
  

42	
  

Overall,	
   these	
   experiments	
   are	
   consistent	
   with	
   the	
   dimer/monomer	
   ratio	
   being	
   kinetically	
  
controlled,	
  with	
  higher	
  concentrations	
  favoring	
  dimer	
  formation.	
  (45)	
  In	
  the	
  meanwhile,	
  we	
  were	
  
investigating	
   the	
   possibility	
   that	
   WT	
   hCRBPII	
   could	
   also	
   form	
   a	
   dimeric	
   species	
   by	
  
overexpressing	
  hCRBPII	
  at	
  the	
  dimer	
  favoring	
  induction	
  temperature	
  of	
  30	
  °C,	
  resulting	
  in	
  both	
  
monomeric	
  and	
  dimeric	
  protein	
  (Figure	
  II-­‐5i).	
  To	
  our	
  knowledge,	
  this	
  was	
  the	
  first	
  example	
  of	
  a	
  
domain	
   swapped	
   system	
   where	
   both	
   monomer	
   and	
   dimer	
   of	
   the	
   wild-­‐type	
   species	
   are	
  
expressed	
  without	
  monomer/dimer	
  exchange	
  (Figure	
  II-­‐8).	
  (45)	
  	
  
	
  
	
  

a	
  

b	
  

	
  
	
  
	
  
	
  
Figure	
   II-­‐8:	
   a.The	
   size	
   exclusion	
   chromatogram	
   for	
   dimer	
   of	
   WT	
   hCRBPII,	
   after	
  
maintaining	
  the	
   protein	
  at	
  room	
  temperature	
  for	
  several	
  days,	
  showing	
  only	
  the	
  dimer	
  
species.	
  b.	
  Gel	
  filtration	
  chromatogram	
  of	
  monomer	
  of	
  WT,	
  showing	
  only	
  the	
  monomer	
  
protein.	
  

	
  Binding	
  assays	
  also	
  show	
  that	
  this	
  dimer	
  has	
  a	
  similar	
  affinity	
  (in	
  nM	
  range,	
  compared	
  
to	
  monomer)	
  for	
  retinol	
  and	
  retinal	
  as	
  well	
  (Figure	
  II-­‐9,	
  method	
  explained	
  in	
  II-­‐3-­‐2	
  section).	
  	
  
	
  
	
  

43	
  

	
  
a

	
  

b

Kd = 5.64 nM

c

	
  
Kd = 180 nM

Kd = 4.36 nM

d
Kd = 110 nM

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

	
  
Figure	
   II-­‐9:	
   	
   Ligand	
   binding,	
   monitored	
   by	
   Tryptophan	
   fluorescence	
   quenching	
   of	
   	
  
monomeric	
  and	
  dimeric	
  hCRBPII.	
  a.	
  Monomer	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinol.	
  b.	
   	
  
Dimer	
   WT	
   hCRBPII	
   with	
   all-­‐trans-­‐retinol.	
   c.	
   Monomer	
   WT	
   hCRBPII	
   with	
   all-­‐trans-­‐ 	
  
retinal.	
  d.	
  Dimer	
  WT	
  hCRBPII	
  with	
  all-­‐trans-­‐retinal.	
  	
  	
  
	
  
	
  
	
  
	
  

II-­‐4	
  Structural	
  Studies	
  
II-­‐4-­‐1	
  Structural	
  analysis	
  reveals	
  an	
  extensive	
  domain	
  swapped	
  dimer	
  
We	
  were	
  able	
  to	
  crystallize	
  and	
  determine	
  several	
  dimer	
  structures	
  to	
  understand	
  the	
  
nature	
  of	
  this	
  new	
  form	
  of	
  hCRBPII.	
  (45)	
  (Figure	
   II-­‐3)	
  In	
  this	
  attempt,	
  structures	
  of	
  dimer	
  mutants	
  
(KLY60W,	
  Y60W	
  and	
  Y60L)	
  and	
  wild	
  type	
  (WT)	
  were	
  obtained	
  respectively	
  (Table	
  II-­‐7).	
  The	
  same	
  
extensive,	
  domain	
  swapped	
  dimer	
  was	
  observed	
  for	
  all	
  four	
  of	
  these	
  structures	
  (Figures	
  II-­‐3,	
  10,	
  

44	
  

11,	
   12a	
   and	
   12b).	
  The	
  domain	
  swapping	
  is	
  extensive,	
  involving	
  residue	
  1-­‐56,	
  which	
  includes	
  3	
  
beta	
  strands	
  and	
  two	
  helices	
  (Figure	
  II-­‐3,	
  12a,	
  12b).	
  	
  	
  
	
  
	
  

aa

Arg58
Trp60

	
  

Phe57

Ser55

	
  
Asn59

	
  

Thr56

	
  

bb

Phe57

Trp60
Asn59

	
  
Ser55

	
  

Thr56
Arg58

	
  
	
  

c

c

	
  
	
  
	
  
	
  
d

Figure	
   II-­‐10:	
   Crystal	
  structure	
  of	
  dimer	
  KLY60W	
  mutant	
  and	
  its	
  monmer	
  components	
  a.	
  
Chain	
   A	
   of	
   KLY60W	
   (where	
   only	
   the	
   Thr56	
   psi	
   angle	
   differs	
   from	
   the	
   closed	
   monomer	
  
form)	
   with	
   the	
   electron	
   density,	
   contoured	
   at	
   1.0	
   σ,	
   of	
   the	
   hinge	
   loop	
   region	
  
encompassing	
   amino	
   acids	
   55-­‐60.	
   	
   Atoms	
   colored	
   by	
   type,	
   N,	
   blue,	
   O,	
   red,	
   C,	
   cyan.	
   b.	
  
Chain	
   B,	
   showing	
   the	
   electron	
   density,	
   contoured	
   at	
   1.0	
   σ,	
   of	
   the	
   hinge	
   loop	
   region	
  
encompassing	
  amino	
   acids	
  55-­‐60.	
  Atoms	
  colored	
  by	
  type,	
   N,	
  blue,	
  O,	
   red,	
  C,	
  cyan.	
   c.	
   The	
  
complete	
   structure	
   of	
   the	
  KLY60W	
  dimer:	
  In	
  chain	
   A	
  (cyan)	
  Trp60	
   is	
  pointing	
   inside	
  the	
  
binding	
  pocket,	
  while	
  in	
  chain	
  B	
  (pink)	
  Trp60	
  is	
  pointing	
  toward	
  the	
  solvent.	
  	
  

45	
  

This	
  corresponds	
  to	
  an	
  83%	
  extent	
  of	
  swapping	
  as	
  defined	
  in	
  the	
  domain	
  swapping	
  3D	
  
knowledge	
   database(http://caps.ncbs.res.in/3dswap/index.html).(83)	
   Each	
   of	
   the	
   two	
   domains	
  
in	
   the	
   dimer	
   are	
   highly	
   similar	
   to	
   that	
   of	
   the	
   monomer	
   hCRBPII	
   structure	
   with	
   RMSD	
   ranging	
  
from	
  0.382-­‐	
  0.397	
  between	
  monomeric	
  hCRBPII	
  and	
  a	
  single	
  domain	
  of	
  the	
  dimer	
  (Figure	
   II-­‐11).	
  
These	
  domain	
  swapped	
  dimer	
  structures	
  have	
  a	
  large	
  rotation	
  about	
  the	
  psi	
  angle	
  of	
  Thr	
  56	
  in	
  
common.	
  	
  	
  

	
  
	
  
	
  
	
  
	
  
Figure	
   II-­‐11:	
   An	
   overlay	
   of	
   the	
   dimeric	
   (cyan)	
   and	
   monomeric	
   (red)	
   KLY60W	
  
mutant	
  structures.	
  Res56	
  for	
  both	
  structures	
  is	
  shown	
  in	
  sticks.	
  The	
  hinge	
  loop	
  
	
  
region	
   in	
   the	
   monomer	
   and	
   dimer	
   is	
   encircled,	
   showing	
   the	
   dramatic	
  
difference	
  in	
  the	
  two	
  structures.	
  
	
  
	
  
While	
  the	
  psi	
  angle	
  of	
  Thr56	
  is	
  about	
  -­‐12°	
  in	
  monomeric	
  hCRBPII,	
  it	
  ranges	
  from	
  132°	
  –	
  
169°	
  in	
  the	
  domain	
  swapped	
  dimers	
  (Figure	
   II-­‐12a)	
   This	
  150°+	
  rotation	
  about	
  the	
  psi	
  angle	
  of	
  
Thr56	
   represents	
   the	
   hinge	
   motion	
   required	
   for	
   domain	
   swapping	
   (Figure	
   II-­‐11).	
   Though	
  
mutation	
  at	
  residue	
  60	
  substantially	
  affects	
  the	
  monomer/dimer	
  ratio,	
  it	
  is	
  not	
  in	
  the	
  hinge	
  loop	
  
region,	
   but	
   is	
   instead	
   3	
   amino	
   acids	
   away,	
   near	
   the	
   middle	
   of	
   β-­‐strand	
   4th.	
   In	
   fact	
   substantial	
  
deviations	
   in	
   torsion	
   angles	
   from	
   the	
   monomer	
   structure	
   can	
   be	
   seen	
   in	
   residues	
   55	
   –	
   60	
   (from	
  
the	
  hinge	
  loop	
  region	
  till	
  even	
  the	
  first	
  3	
  amino	
  acids	
  of	
  the	
  4th	
  beta	
  sheet).	
  	
  

46	
  

a	
  

b	
  

	
  
	
  
	
  
	
  
	
  

c	
  

	
  

d	
  

	
  
	
  
	
  
	
  
	
  

e	
  

	
  
	
  
	
  
	
  
	
   Figure	
   II-­‐12:	
   a-­‐e:	
   Torsion	
   angle	
   differences	
   between	
   monomer	
   and	
   dimer	
  
differences	
  of	
  Phi	
  angle	
  are	
  in	
  blue	
  and	
  psi	
  angle	
  differences	
  are	
  in	
  red	
  in	
  each	
  
	
   case.	
  	
  a.	
  Thr56.	
  b.	
  Phe57.	
  c.	
  Arg58.	
  d.	
  Asn59.	
  e.	
  Res60.	
  	
  	
  
	
   	
  
The	
  domain	
  swapping	
  orients	
  the	
  opening	
  of	
  the	
  two	
  binding	
  sites	
  to	
  face	
  one	
  another,	
  
creating	
   a	
   continuous	
   and	
   large	
   internal	
   cavity	
   that	
   stretches	
   from	
   one	
   side	
   of	
   the	
   dimer	
   to	
   the	
  
other	
  (nearly	
  40	
  Å	
  in	
  length,	
  (please	
  refer	
  to	
  section	
  II-­‐6).	
  A	
  search	
  of	
  the	
  various	
  structural	
  fold	
  
databases	
  (SCOP	
  1.75,	
  CATH	
  3.5)	
  reveals	
  no	
  other	
  protein	
  of	
  similar	
  architecture.(84)	
  

47	
  

II-­‐4-­‐2	
  Symmetry	
  VS	
  asymmetry	
  	
  
Although	
   all	
   four-­‐dimer	
   structures	
   shared	
   the	
   same	
   domain	
   swapped	
   interface,	
   there	
   are	
  
substantial	
   differences	
   in	
   their	
   structures.	
   As	
   crystallographers	
   and	
   structural	
   biologists,	
   we	
  
believed	
  that	
  the	
  structural	
  details	
  of	
  the	
  DS	
  dimers	
  are	
  key	
  to	
  understanding	
  the	
  mechanism	
  of	
  
domain	
  swapping.	
  	
  
a

	
  
	
  

b

	
  

	
  
	
  
	
  

c

d

	
  
	
  
	
  
Figure	
   II-­‐13:	
   The	
   differences	
   in	
   the	
   asymmetric	
   and	
   symmetric	
   structures	
   of	
   hCRBPII	
  
dimers.	
  Res60	
   in	
   all	
   structures	
  is	
   shown	
   in	
   sticks	
  a.	
   The	
   asymmetric	
  structures:	
   The	
   overlay	
  
structure	
   of	
   domain	
   swapped	
   dimer	
   KLY60W	
   (cyan)	
   and	
   Y60W	
   (pink).	
   In	
   chain	
   A	
   dimer	
  
KLY60W,	
   Trp60	
   is	
   pointing	
   inside	
   the	
   binding	
   pocket.	
   b.	
   The	
   symmetric	
   structures.	
   The	
  
overlay	
   of	
   WT	
   hCRBPII	
   (purple)	
   and	
   Y60L	
   (yellow)	
   domain	
   swapped	
   dimer.	
   c.	
   Comparing	
  
asymmetric	
  (dimer	
  KLY60W	
  in	
  cyan)	
  and	
  symmetric	
  dimer	
  (dimer	
  Y60L	
  in	
  yellow)	
  by	
  looking	
  
down	
  the	
  two-­‐fold	
   axis	
   and	
   it	
   shows	
  the	
   different	
   position	
   of	
   helices	
   in	
   N-­‐terminus.	
   d.	
   One	
  
chain	
  of	
  dimer	
  Y60L	
  (yellow)	
  and	
  both	
  chains	
  of	
  asymmetric	
  dimer	
  Y60W	
  (chain	
  A	
  in	
  cyan	
  
and	
  chain	
  B	
  in	
  red)	
  are	
  overlaid	
  at	
  their	
  C-­‐terminus.	
  
We	
   found	
   that	
   hCRBPII	
   DS	
   dimers	
   can	
   be	
   classified	
   as	
   either	
   symmetric	
   (WT	
   hCRBPII,	
  
Y60L	
   and	
   other	
   mutants	
   described	
   in	
   II-­‐6	
   section),	
   where	
   the	
   two	
   chains	
   which	
   make	
   the	
   dimer	
  
are	
  essentially	
  identical,	
  or	
  asymmetric	
  (KLY60W,	
  Y60W	
  and	
  holo	
  DS	
  dimers	
  described	
  in	
  II-­‐6),	
  
where	
  the	
  two	
  chains	
  are	
  distinct	
  from	
  one	
  another	
  (Figure	
   II-­‐13). (45)	
  	
  As	
  shown	
  in	
  Figure	
   II-­‐13b	
  
48	
  

the	
  two	
  symmetric	
  structures	
  (WT	
  and	
  Y60L)	
  are	
  similar,	
  with	
  the	
  relative	
  orientation	
  of	
  the	
  two	
  
domains	
   in	
   the	
   dimer	
   essentially	
   identical.	
   	
   However,	
   there	
   are	
   significant	
   differences	
   in	
   the	
  
torsion	
  angles	
  around	
  residue	
  60,	
  with	
  an	
  associated	
  repositioning	
  of	
  the	
  side	
  chain.	
  	
  Tyr60	
  is	
  
buried	
  inside	
  of	
  the	
  protein,	
  making	
  a	
  hydrogen	
  bond	
  with	
  Glu72	
  in	
  both	
  the	
  WT	
  monomer	
  and	
  
dimer,	
  while	
  Leu60	
  is	
  flipped	
  out	
  of	
  the	
  inside	
  of	
  the	
  protein	
  and	
  solvent	
  exposed	
  in	
  the	
  Y60L	
  
dimer	
  (Figure	
  II-­‐14a).	
  	
  
This	
   “flipped	
   out”	
   conformation	
   is	
   the	
   cause	
   of	
   the	
   residue	
   59	
   phi	
   angle	
   large	
   deviation,	
  
relative	
   to	
   its	
   monomer	
   value	
   (about	
   -­‐121°	
   in	
   the	
   monomer	
   versus	
   -­‐80°	
   in	
   the	
   Y60L	
   dimer)	
  
(Figure	
  II-­‐14c)	
  In	
  contrast,	
  the	
  phi	
  torsion	
  angle	
  of	
  Asn59	
  in	
  the	
  WT	
  dimer	
  is	
  -­‐142°	
  (similar	
  to	
  its	
  
monomer	
   value),	
   which	
   leads	
   to	
   the	
   “flipped	
   in”	
   conformation	
   (Figures	
   II-­‐14a	
   and	
   14b).	
   	
   To	
  
balance	
   the	
   resulting	
   change	
   in	
   trajectory	
   of	
   the	
   main	
   chain,	
   the	
   Asn59	
   psi	
   angle	
   is	
   radically	
  
rotated	
   relative	
   to	
   all	
   of	
   the	
   other	
   structures	
   (-­‐142°	
   in	
   the	
   WT	
   dimer	
   versus	
   161°	
   in	
   the	
  
monomer	
  and	
  165	
  in	
  the	
  dimer	
  Y60L,	
  see	
  Figures	
  II-­‐b),	
  resulting	
  in	
  a	
  substantial	
  Ramachandran	
  
outlier	
  at	
  Asn59.	
  Nonetheless	
  it	
  results	
  in	
  the	
  main	
  chains	
  of	
  the	
  two	
  symmetric	
  structures	
  once	
  
again	
  closely	
  tracking	
  each	
  other	
  (Figure	
  II-­‐14a).	
  
	
  
	
  

49	
  

a 	
  

	
  
2.6 Å

	
  

Tyr60
Glu72

Leu60

	
  

Asn59

	
  
b

	
  

c

	
  
	
  

Glu72
Glu72

2.6 Å
Tyr60

	
  
	
  
	
  

Leu60

Asn59

ÎŚ = -142Âş
Ψ = -142º
(161Âş in monomer)

Asn59
ÎŚ = -80Âş
(-121Âş in monomer)
Ψ = 165º

Figure	
   II-­‐14.	
   Torsion	
   angle	
   deviations	
   outside	
   the	
   hinge-­‐loop	
   region	
   define	
   the	
  
	
  
relative	
  orientation	
  of	
  the	
  two	
  domains	
  of	
  the	
  dimer,	
  with	
  the	
  highest	
  deviation	
  
seen	
  in	
  Asn59.	
  a.	
  	
  The	
  WT	
  (purple)	
  and	
  Y60L	
  (yellow)	
  dimers	
  are	
  overlaid.	
  	
  Inset,	
  
	
  
the	
   hydrogen	
   bond	
   between	
   Tyr60	
   and	
   Glu72	
   compared	
   to	
   the	
   “flipped	
   out”	
  
conformation	
   of	
  Leu60.	
  b.	
  The	
   critical	
  Tyr60	
  and	
  Asn59	
  region	
   in	
  the	
  WT	
   hCRBPII	
  
	
  
dimer,	
   showing	
   the	
   key	
   phi/psi	
   angles.	
   c.	
   The	
   same	
   region	
   in	
   the	
   Y60L	
   dimer,	
  
showing	
  the	
  “flipped	
  out”	
  Leu60	
  and	
  the	
  key	
  phi/psi	
  angles.	
  	
  Comparison	
  of	
  the	
  
	
  
two	
  shows	
  how	
  the	
  large	
  difference	
  in	
  the	
  phi	
  angle	
  is	
  compensated	
  for	
  in	
  the	
  WT	
  
N59	
  psi	
  angle,	
  keeping	
  the	
  main	
  chain	
  of	
  the	
  two	
  on	
  a	
  similar	
  trajectory.	
  	
  
	
  
In	
   contrast	
   to	
   the	
   symmetric	
   dimers,	
   asymmetric	
   dimer	
   (i.e.	
   Y60W	
   and	
   KLY60W)	
  
structures	
   are	
   similar	
   to	
   one	
   another	
   and	
   asymmetric	
   (Figures	
   II-­‐13a).	
   However,	
   there	
   is	
   a	
  
significant	
  deviation	
  in	
  the	
  position	
  of	
  the	
  N-­‐terminal	
  region	
  when	
  the	
  C-­‐terminal	
  domains	
  are	
  
overlaid	
  (Figures	
   II-­‐13c	
   and	
   13d).	
  Basically,	
  in	
  asymmetric	
  dimers,	
  there	
  is	
  a	
  large	
  deviation	
  in	
  
the	
   relative	
   orientation	
   of	
   the	
   N-­‐	
   and	
   C-­‐terminal	
   regions	
   in	
   the	
   two-­‐polypeptide	
   chains	
  
(subunits)	
  that	
  make	
  the	
  dimer.	
  (Figure	
  II-­‐13d) (45)	
  

50	
  

	
  
V62

	
  

R58
W60
F57

	
  
D63

N59

	
  

D61
T56

	
  
V62

	
  

N59

W60

F57

	
  
D63

D61

	
  

R58

T56

	
  
Figure	
  
II-­‐15:	
   Residues	
   56-­‐63	
   in	
   the	
   chains	
   A	
   (green,	
   top)	
   and	
   B	
   (magenta,	
  
	
  
bottom)	
  of	
  Y60W	
  hCRBPII.	
  
	
  	
  
This	
  positional	
  deviation	
  is	
  as	
  large	
  as	
  14	
  Å	
  between	
  the	
  two	
  conformations	
  of	
  the	
  DS	
  
dimer.	
  	
  This	
  represents	
  a	
  radical	
  change	
  in	
  the	
  relative	
  orientation	
  of	
  the	
  two	
  domains	
  relative	
  
to	
  the	
  symmetric	
  dimer.	
  	
  	
  
The	
   source	
   of	
   this	
   large	
   deviation	
   can	
   be	
   seen	
   in	
   the	
   torsion	
   angles	
   of	
   the	
   residues	
  
between	
  56-­‐60	
  (Figures	
  II-­‐12).	
  	
  Where	
  in	
  chain	
  A	
  of	
  both	
  KLY60W	
  and	
  Y60W	
  dimers,	
  all	
  of	
  the	
  
torsion	
  angles,	
  except	
  the	
  psi	
  angle	
  of	
  Thr56,	
  are	
  similar	
  to	
  those	
  of	
  a	
  monomer.	
  In	
  chain	
  B	
  a	
  
number	
  of	
  other	
  torsion	
  angles,	
  namely	
  the	
  psi	
  angles	
  of	
  residues	
  Thr56	
  and	
  Phe57,	
  and	
  the	
  phi	
  
angle	
   of	
   Phe57,	
   deviate	
   from	
   their	
   monomer	
   values	
   (Figures	
   II-­‐12a	
   and	
   12b).	
   	
   Presumably,	
  
these	
  differences	
  in	
  torsion	
  angles	
  are	
  required	
  to	
  orient	
  the	
  two	
  halves	
  of	
  the	
  chain	
  properly	
  
for	
  dimer	
  formation.	
  

51	
  

Interestingly,	
  the	
  A	
  subunit	
  of	
  the	
  KLY60W	
  and	
  Y60W	
  mutants	
  is	
  the	
  most	
  similar	
  to	
  the	
  
monomer.	
  It	
  has	
  phi/psi	
  angles	
  almost	
  identical	
  to	
  those	
  of	
  the	
  monomer,	
  except	
  the	
  psi	
  angle	
  
of	
  hinge	
  residue	
  T56,	
  which	
  is	
  rotated	
  by	
  almost	
  180	
  degrees	
  in	
  all	
  the	
  DS	
  dimers.	
  Phe57	
  and	
  
Arg58	
  side	
  chains	
  are	
  on	
  the	
  same	
  side	
  of	
  the	
  strand,	
  similar	
  to	
  the	
  conformation	
  seen	
  in	
  the	
  
loop	
  connecting	
  the	
  two	
  strands	
  in	
  the	
  monomer	
   (Figure	
   II-­‐12c	
   and	
   15).	
   However,	
  the	
  phi/psi	
  
angles	
  of	
  the	
  B	
  subunit	
  had	
  other	
  significant	
  differences	
  with	
  monomer,	
  especially	
  the	
  phi	
  angle	
  
of	
   residue	
   59	
   (Figure	
   II-­‐12d),	
   which	
   resulted	
   in	
   the	
   flipping	
   of	
   Trp60	
   from	
   inside	
   the	
   binding	
  
cavity	
  to	
  the	
  outside	
  and	
  Asn59	
  from	
  outside	
  to	
  inside	
  of	
  the	
  cavity	
  (relative	
  to	
  the	
  monomer)	
  
(Figure	
   II-­‐15).	
   This	
  places	
  both	
  residues	
  60	
  and	
  61	
  outside	
  the	
  binding	
  cavity,	
  which	
  is	
  necessary	
  
to	
   re-­‐phase	
   the	
   strand	
   (explained	
   in	
   the	
   next	
   section,	
   II-­‐4-­‐2).	
   The	
   asymmetry	
   of	
   the	
   Y60W	
  
mutant	
  dimers	
  tells	
  us	
  that	
  the	
  “canonical”	
  DS	
  dimer,	
  where	
  the	
  only	
  torsion	
  angle	
  change	
  is	
  in	
  
the	
   hinge	
   residue,	
   is	
   not	
   possible	
   because	
   the	
   relative	
   orientations	
   of	
   the	
   N-­‐	
   and	
   C-­‐terminal	
  
regions	
   of	
   each	
   domains	
   are	
   incompatible	
   with	
   dimer	
   formation;	
   thus	
   subunit	
   B	
   must	
   adjust	
  
itself,	
   via	
   the	
   torsion	
   angles	
   along	
   the	
   connecting	
   strand,	
   to	
   adopt	
   a	
   conformation	
   consistent	
  
with	
  dimer	
  formation	
  to	
  subunit	
  A.(45)	
  (Figure	
  II-­‐16)

52	
  

	
  
	
  
	
  

Figure	
   II-­‐16:	
   Overview	
  of	
  dimer	
  formation	
  in	
  asymmetric	
  structures:	
  Two	
  chains	
  of	
  
Y60W	
  dimer	
  are	
  labeled	
  in	
  the	
  picture.	
  The	
  same	
  chain	
  (molecule	
  A	
  and/or	
  B)	
  in	
  
this	
   asymmetric	
   dimers	
   cannot	
   form	
   dimers	
   since	
   they	
   clash.	
   However,	
   the	
  
interaction	
   of	
   different	
   chain	
   with	
   each	
   other	
   (molecule	
   A	
   and	
   B)	
   could	
   lead	
   to	
  
proper	
  DS	
  dimer	
  formation	
  with	
  the	
  different	
  orientation.	
  

II-­‐4-­‐3	
  Phase	
  Relationship	
  
	
  

In	
  all	
  cases	
  domain	
  swapping	
  is	
  the	
  result	
  of	
  the	
  straightened	
  connecting	
  loop	
  between	
  

beta	
   strands	
   3	
   	
   and	
   4	
   of	
   the	
   monomer,	
   resulting	
   in	
   this	
   loop	
   and	
   beta	
   strands	
   3rd	
   and	
   4th	
  
becoming	
  a	
  single	
  beta	
  strand	
  stretching	
  the	
  length	
  of	
  the	
  dimer	
  (Figure	
  II-­‐3).	
  	
  
	
  
	
  

53	
  

	
  

N59

D63

D61

= Carbonyl C
=ÎąC
= Nitrogen

V62

Y60

R58

= β Sheet, Side Chain In

T51

T53

S55

K50

K52

T54

	
  

F57
T56

= β Sheet, Side Chain Out
= Loop

	
  

Monomer

	
  
T51

	
  

K50

T53

K52

S55

T54

F57

T56

V62

N59

R58

Res
60

D61

D63

Dimer

	
  
Figure	
   II-­‐17:	
   Schematic	
   representation	
   of	
   “phase	
   relationship”	
   in	
   DS	
  
dimerization.	
  
Top:	
  monomer,	
  bottom:	
  dimer	
  with	
  residue	
  60	
  as	
  mutant.	
  
	
  
In	
  order	
  to	
  have	
  a	
  DS	
  dimer,	
  with	
  a	
  perfect	
  stretched	
  beta	
  sheet	
  in	
  that	
  region,	
  eventually	
  we	
  
will	
  have	
  “phase	
  problem”	
  because,	
  in	
  monomer	
  while	
  the	
  odd	
  side	
  chains	
  of	
  beta	
  strand	
  3	
  face	
  
inside	
   the	
   binding	
   pocket,	
   the	
   even	
   side	
   chains	
   of	
   beta	
   strand	
   4	
   face	
   inside.	
   With	
   an	
   even	
  
number	
  of	
  residues	
  in	
  the	
  loop,	
  this	
  means	
  that	
  the	
  single	
  beta	
  strand	
  formed	
  in	
  the	
  DS	
  dimer	
  
must,	
  at	
  some	
  point,	
  re-­‐phase	
  the	
  strand,	
  which	
  would	
  then	
  put	
  the	
  even	
  side	
  chains	
  of	
  the	
  C-­‐
terminal	
  half	
  of	
  the	
  strand	
  “in	
  phase”	
  with	
  the	
  odd	
  numbered	
  side	
  chains	
  of	
  the	
  first	
  half.	
  	
  
The	
  way	
  that	
  mutants	
  (i.e.	
  Y60L	
  and	
  Y60W)	
  solve	
  the	
  phase	
  problem	
  might	
  be	
  the	
  key	
  
point	
   in	
   the	
   domain	
   swapping	
   mechanism.	
   Conformations	
   of	
   residues	
   Y60	
   and	
   D61	
   are	
  
important	
   to	
   consider	
   as	
   well.	
   In	
   Y60L,	
   Y60W	
   (chain	
   B),	
   KLY60W	
   (chain	
   B)	
   mutant	
   domain	
  
swapped	
  dimers	
  of	
  hCRBPII,	
  these	
  two	
  residues	
  are	
  pointed	
  toward	
  the	
  solvent,	
  which	
  re-­‐phase	
  
the	
  strand	
  in	
  order	
  to	
  fix	
  the	
  phase	
  problem	
  (figure	
   II-­‐17).	
  However,	
  in	
  wild	
  type	
  hCRBPII,	
  which	
  
gave	
  us	
  mostly	
  monomer,	
  residue	
  Y60	
  is	
  inside	
  the	
  binding	
  pocket.	
  This	
  may	
  suggest	
  one	
  of	
  the	
  

54	
  

significant	
  points	
  in	
  the	
  mechanism	
  for	
  dimerization	
  in	
  hCRBPII.

II-­‐5	
  possible	
  mechanism	
  for	
  domain	
  swapping	
  in	
  hCRBPII	
  	
  
With	
   the	
   existence	
   of	
   the	
   WT	
   domain	
   swapped	
   dimer	
   and	
   the	
   fact	
   that	
   there	
   is	
   no	
  
Interconversion	
  between	
  monomer	
  and	
  dimer,	
  indicating	
  each	
  as	
  a	
  unique	
  fold	
  of	
  this	
  protein.	
  	
  	
  
Therefore,	
  we	
  proposed	
  a	
  mechanism	
  for	
  the	
  hCRBPII	
  folding	
  pathway,	
  different	
  from	
  what	
  has	
  
been	
   seen	
   for	
   other	
   iLBPs	
   earlier(72).	
   	
   We	
   suggested	
   that	
   the	
   two	
   halves	
   of	
   hCRBPII	
   (the	
   N-­‐
terminal	
  and	
  C-­‐terminal	
  halves)	
  are	
  capable	
  of	
  at	
  least	
  partially	
  folding	
  independently,	
  initially	
  
as	
   an	
   extended	
   “open	
   monomer.”	
   	
   The	
   dimer/monomer	
   ratio	
   would	
   then	
   depend	
   on	
   the	
  
relative	
   rates	
   of	
   dimerization	
   of	
   the	
   open	
   monomers,	
   versus	
   rotation	
   of	
   the	
   n-­‐	
   and	
   c-­‐termini	
  
together	
  to	
  form	
  the	
  “closed	
  monomer”. 	
  
Our	
  data	
  shows	
  that	
  three	
  things	
  are	
  required	
  for	
  domain	
  swapping	
  in	
  this	
  system,	
  the	
  
large	
  hinge	
  motion	
  at	
  Thr56,	
  subsequent	
  proper	
  orientation	
  of	
  the	
  n-­‐	
  and	
  c-­‐termini	
  to	
  adopt	
  a	
  
conformation	
  consistent	
  with	
  dimer	
  formation	
  and	
  re-­‐phasing	
  of	
  the	
  connecting	
  strand	
  with	
  the	
  
last	
  two	
  strongly	
  correlated.	
  (Figure	
  II-­‐18)	
  	
  
	
  
	
  
	
  
	
  

55	
  

T56

F57
HOOC

	
  

NH2

	
  
	
  

H2N

COOH

Folded Monomer

	
  
NH2

	
  

NH2
H2N

COOH

Domain Swapped Dimer

	
  
T56

	
  

T56
F57

F57

	
  
	
  
	
  
	
  
	
  

HOOC

Open Monomer

COOH

Rephased Open Monomer

Figure	
   II-­‐18:	
   DS	
   dimerization	
   requires:	
   1.	
   Rotation	
   about	
   Thr56	
   psi.	
   2.	
  
Orientation	
   of	
   the	
   two	
   halves	
   to	
   accommodate	
   dimerization.	
   	
   3.	
  
Rephasing	
  of	
  the	
  connecting	
  strand.	
  

As	
  one	
  would	
  expect,	
  this	
  is	
  a	
  rate-­‐governed	
  process,	
  thus	
  factors	
  such	
  as	
  temperature	
  
and	
  concentration	
  should	
  affect	
  the	
  outcome.	
  In	
  all	
  the	
  symmetric	
  dimers	
  thus	
  far	
  structurally	
  
characterized	
   virtually	
   the	
   same	
   dimer	
   is	
   seen,	
   where	
   both	
   subunits	
   of	
   the	
   dimer	
   have	
   very	
  
similar	
   relative	
   orientations	
   of	
   their	
   n-­‐and	
   c-­‐termini,	
   with	
   Asn59	
   flipped	
   inside	
   the	
   binding	
  
cavity,	
  which	
  is	
  “in	
  phase”	
  with	
  the	
  n-­‐terminal	
  side	
  of	
  the	
  connecting	
  beta	
  strand.	
  (Figure	
  II-­‐21)	
  
However,	
   reaching	
   this	
   symmetric	
   dimer	
   arrangement	
   is	
   accomplished	
   in	
   different	
   ways.	
  
Flipping	
  out	
  of	
  the	
  residue	
  at	
  position	
  60,	
  in	
  both	
  the	
  Y60L	
  symmetric	
  dimer	
  and	
  molecule	
  B	
  of	
  
both	
   asymmetric	
   dimers	
   (Figures	
   II-­‐14a	
   and	
   15)	
   is	
   one	
   low	
   energy	
   pathway	
   for	
   a	
   “dimer	
  
friendly”	
  orientation.	
  In	
  contrast	
  WT	
  hCRBPII	
  is	
  “spring	
  loaded”	
  against	
  this	
  conformation	
  by	
  the	
  

56	
  

hydrogen	
  bond	
  made	
  between	
  Glu72	
  and	
  Tyr60,	
  which	
  serves	
  to	
  hold	
  Tyr60	
  in	
  the	
  “flipped	
  in”	
  
position	
  (Figures	
  II-­‐14b).	
  	
  	
  
Formation	
   of	
   the	
   WT	
   symmetric	
   dimer	
   then	
   requires	
   an	
   almost	
   90°	
   rotation	
   of	
   the	
  
Asn59	
  psi	
  angle	
  resulting	
  in	
  a	
  substantial	
  Ramachandran	
  outlier	
  (Figure	
  II-­‐12d).	
  (“Twist	
  against	
  
the	
   spring”	
   situation).	
   	
   This	
   gave	
   rise	
   to	
   the	
   small	
   quantities	
   of	
   WT	
   dimer	
   and	
   its	
   relative	
  
instability.	
  Tyr	
  at	
  position	
  60	
  is	
  resistant	
  to	
  flipping	
  out	
  due	
  to	
  the	
  hydrogen	
  bond	
  with	
  Glu72,	
  
which	
   “spring	
   loaded”	
   it	
   against	
   dimerization,	
   (Figure	
   II-­‐19)	
   and	
   therefore	
   requiring	
   an	
  
unfavorable	
   psi	
   angle	
   in	
   Asn59	
   to	
   achieve	
   the	
   relative	
   orientation	
   required	
   for	
   dimerization	
  
(Figure	
  II-­‐	
  12d).	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   II-­‐19:	
   DS	
   dimer	
   vs.	
   monomer	
   formation	
   in	
   WT	
  
hCRBP	
  II.	
  
	
  

In	
   none	
   of	
   the	
   other	
   DS	
   dimer	
   structures	
   we	
   see	
   Ramachandran	
   outliers	
   in	
   the	
  
connecting	
   strand	
   region,	
   indicating	
   since	
   Tyr60	
   is	
   inside	
   the	
   binding	
   pocket	
   and	
   making	
   its	
  
hydrogen	
   bond	
   with	
   Glu72	
   results	
   in	
   a	
   relatively	
   high	
   energetic	
   penalty	
   for	
   dimerization,	
   due	
  

57	
  

both	
  subunits	
  to	
  the	
  need	
  to	
  rephase	
  the	
  sidechains	
  on	
  the	
  connecting	
  strand,	
  and	
  to	
  allow	
  the	
  
n-­‐	
  and	
  c-­‐termini	
  to	
  adopt	
  a	
  conformation	
  consistent	
  with	
  dimerization.	
  	
  
Based	
   on	
   this	
   hypothesis	
   mutation	
   of	
   E72	
   to	
   a	
   residue	
   incompetent	
   for	
   hydrogen	
  
bonding	
   (a	
   small	
   hydrophobic	
   residue)	
   will	
   “loosen	
   the	
   spring”	
   resulting	
   in	
   an	
   increase	
   in	
   dimer	
  
formation.	
  (Figure	
  II-­‐19)	
  	
  	
  

II-­‐5-­‐1	
  Dimerization	
  and	
  structure	
  of	
  the	
  E72A	
  mutant	
  
As	
   a	
   first	
   step	
   to	
   confirm	
   and	
   test	
   the	
   abovementioned	
   hypothesis,	
   the	
   E72A	
   mutant	
   was	
  
created,	
   expressed	
   and	
   purified.	
   	
   As	
   predicted	
   this	
   mutation	
   led	
   to	
   a	
   dramatic	
   increase	
   in	
  
dimerization	
   relative	
   to	
   WT	
   hCRBPII.	
   	
   (E72A	
   mutant	
   produced	
   50%	
   dimer	
   vs.	
   only10%	
   in	
   WT)	
  
(Table	
  II-­‐9,	
  Figure	
  II-­‐20).	
  	
  The	
  E72A	
  mutant	
  was	
  then	
  crystallized	
  and	
  its	
  structure	
  determined.	
  	
  
	
  
a

	
  

b

	
  
	
  
	
  
	
  
	
  

Figure	
   II-­‐20:	
  Size	
  exclusion	
   chromatography	
  of	
  low	
  salt	
  and	
   high	
   salt	
   (80	
   mM	
   and	
  
160	
   mM	
   NaCl,)	
  E72A	
   hCRBPII	
   respectively,	
   expressed	
   at	
  25°C	
  (since	
  30°C	
  expressed	
  
produced	
  virtually	
  no	
  soluble	
  protein).	
  
	
  

Although	
  a	
  crystallographic	
  two-­‐fold	
  axis	
  does	
  not	
  relate	
  the	
  two	
  monomers	
  in	
  

this	
  dimer,	
  the	
  relative	
  orientation	
  of	
  the	
  N-­‐	
  and	
  C-­‐terminal	
  halves	
  of	
  the	
  protein	
  are	
  essentially	
  
identical	
   to	
   that	
   of	
   the	
   symmetric	
   dimers	
   (Figure	
   II-­‐21C).	
   There	
   are	
   deviations	
   in	
   its	
   torsion	
  

58	
  

angles	
  at	
  position	
  57	
  (phi	
  angles	
  of	
  -­‐85°	
  versus	
  -­‐133)	
  and	
  most	
  notably,	
  position	
  60	
  (phi	
  angles	
  
of	
   35°	
   versus	
   -­‐72°	
   and	
   psi	
   angles	
   of	
   92°	
   versus	
   127)	
   (Figure	
  II-­‐12b	
  and	
  e).	
   	
   The	
   result	
   is	
   that	
   Tyr	
  
60	
   is	
   in	
   the	
   “flipped	
   out”	
   conformation,	
   similar	
   to	
   that	
   seen	
   in	
   both	
   the	
   Y60L	
   dimer	
   and	
   no	
  
Ramachandran	
   outliers	
   in	
   the	
   connecting	
   strand	
   residues	
   (Figure	
   II-­‐21a),	
   thus	
   inducing	
   the	
  
flexibility	
   between	
   the	
   n-­‐	
   and	
   c-­‐terminal	
   domains	
   (compare	
   to	
   WT,	
   Figure	
   II-­‐21b),	
   favoring	
  
dimer	
  formation,	
  and	
  suggesting	
  that	
  the	
  hydrogen	
  bond	
  between	
  Tyr60	
  and	
  Glu72	
  is	
  there	
  to	
  
repress	
  dimer	
  formation	
  and	
  consistent	
  with	
  the	
  hypothesis	
  that	
  the	
  interaction	
  between	
  Y60	
  
and	
   E72	
   in	
   the	
   open	
   monomer	
   folding	
   intermediate	
   effectively	
   holds	
   the	
   two	
   domains	
   in	
   a	
  
conformation	
   unfavorable	
   to	
   dimerization,	
   promoting	
   the	
   formation	
   of	
   the	
   physiologically	
  
relevant	
  monomeric	
  species.	
  	
  This	
  suggests	
  that	
  the	
  n-­‐	
  and	
  c-­‐terminal	
  domains	
  are	
  capable	
  of	
  at	
  
least	
  partially	
  folding	
  in	
  isolation,	
  and	
  would	
  represent	
  the	
  “foldons”(85-­‐89)	
  for	
  hCRBPII.	
  (Figure	
  II-­‐
18	
  and	
  19)	
  Note	
  that	
  Tyr60	
  is	
  not	
  conserved	
  in	
  the	
  iLBP	
  family,	
  and	
  is	
  found	
  only	
  in	
  CRBP’s.	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

59	
  

a

	
  

a

	
  

Glu72
Ala72

	
  

Tyr60

Leu60
Asn59

	
  
	
  

b
b

	
  

Ala72
Tyr60

	
  

Asn59

Tyr60

Glu72
2.6 Å

	
  
	
  

c

c

	
  
	
  
	
  
	
  
Figure	
   II-­‐21:	
   a.	
   Overlay	
   of	
   chain	
   A	
   of	
   the	
   E72A	
   dimer	
   (green)	
   and	
   the	
   Y60L	
   dimer	
   	
  
(yellow).	
  Note	
  that	
  both	
  show	
  residue	
  60	
  pointing	
  toward	
  the	
  solvent.	
  b.	
  Overlay	
  of	
  
the	
  E72A	
  chain	
   A	
  (blue)	
  and	
  WT	
  (green)	
   dimers.	
   Inset	
  shows	
  that	
  without	
   the	
  Glu	
  72	
  
	
  
hydrogen	
   bond	
  donor,	
  residue	
  60	
  is	
   free	
  to	
  flip	
  out,	
  leading	
   to	
   a	
  more	
  relaxed	
  dimer	
  
structure.	
   c.	
   Two	
   chains	
   of	
   the	
   E72A	
   dimer	
   (green	
   and	
   grey)	
   are	
   overlaid,	
   showing	
  
	
  
their	
  strong	
  similarity.	
  
	
  

II-­‐5-­‐2	
  Study	
  the	
  folding	
  pathway	
  	
  
As	
   mentioned	
   above,	
   we	
   proposed	
   that	
   the	
   N	
   and	
   C	
   terminus	
   of	
   the	
   protein	
   might	
   fold	
  
independently	
   	
   (at	
   least	
   partially)	
   to	
   form	
   an	
   open	
   monomer	
   folding	
   intermediate.	
   Then,	
   the	
  
ratio	
   of	
   monomer/dimer	
   would	
   depend	
   on	
   the	
   relative	
   rates	
   of	
   dimerization	
   of	
   open	
  
monomers,	
  versus	
  rotation	
  of	
  the	
  N-­‐	
  and	
  C-­‐termini	
  together	
  to	
  form	
  the	
  “closed	
  monomer”.	
  It	
  

60	
  

was	
   seen	
   before	
   that	
   this	
   is	
   concentration	
   dependent	
   (Table	
   II-­‐1)	
   and	
   would	
   depend	
   on	
   the	
  
energetics	
  of	
  proper	
  orientation	
  for	
  dimer	
  formation.	
  	
  
In	
  this	
  Scenario,	
  the	
  DS	
  dimer	
  would	
  then	
  represent	
  the	
  kineticly	
  trapped	
  folding	
  intermediate	
  
on	
  the	
  native	
  folding	
  pathway	
  of	
  hCRBPII. One	
  of	
  the	
  early	
  steps	
  to	
  detect	
  and	
  study	
  the	
  folding	
  
intermediates	
   is	
   to	
   monitor	
   the	
   protein	
   at	
   different	
   denaturant	
   concentrations,	
   with	
   the	
  
assumption	
   that	
   the	
   structures	
   formed	
   at	
   intermediate	
   denaturant	
   concentrations	
   will	
   reflect	
  
the	
   intermediates	
   formed	
   in	
   the	
   folding	
   pathway.(90,	
   91)	
   To	
   this	
   end,	
   we	
   tried	
   to	
   probe	
   the	
  
structure	
  of	
  WT-­‐hCRBPII	
  over	
  a	
  range	
  of	
  denaturant	
  concentrations	
  (Gd-­‐HCl	
  in	
  this	
  case)	
  using	
  
both	
   Circular	
   dichroism	
   spectroscopy	
   and	
   tryptophan	
   fluorescence	
   spectroscopy.(1)	
   (Figure	
   II-­‐
22)	
   We	
   initially	
   were	
   interested	
   to	
   plot	
   the	
   change	
   in	
   ellipticity	
   versus	
   denaturant,	
   (Figure	
   II-­‐
22c)	
  to	
  compare	
  it	
  to	
  a	
  two-­‐state	
  (N	
  ⇄	
  U)	
  and	
  three	
  state	
  (N	
  ⇄	
  (1/n)In	
  ⇄	
  U	
  or	
  even	
  higher	
  order	
  
model)	
  which	
  would	
  suggest	
  the	
  presence	
  of	
  intermediates	
  (excluded	
  the	
  two-­‐state	
  model)	
  in	
  
the	
   folding	
   pathway(92).	
   It	
   should	
   be	
   also	
   considered	
   that	
   the	
   two	
   halves	
   of	
   the	
   protein	
   may	
  
have	
   significantly	
   different	
   stabilities.	
   After	
   performing	
   these	
   series	
   of	
   experiments,	
   the	
  
helix/strand	
   ratio	
   was	
   calculated	
   from	
   the	
   CD	
   spectra	
   at	
   each	
   individual	
   denaturant	
  
concentration	
   and	
   compared	
   to	
   the	
   secondary	
   structure	
   of	
   the	
   native	
   and	
   predicted	
  
intermediate	
  (which	
  almost	
  consists	
  of	
  the	
  same	
  percentages	
  of	
  helix/strands	
  as	
  native)	
  (Figure	
  
II-­‐19	
  and	
  Table	
  II-­‐8).(1)	
  Also	
  the	
  concentration	
  dependence	
  may	
  suggest	
  that	
  dimer	
  association	
  
occurs	
   early	
   in	
   the	
   protein-­‐folding	
   pathway.	
   	
   Based	
   on	
   the	
   result	
   from	
   our	
   CD	
   spectra	
   and	
  
calculated	
  percentages	
  of	
  helix	
  and	
  beta	
  sheet	
  at	
  different	
  Gd.Hcl	
  concentrations,	
  helixes	
  may	
  
be	
  more	
  stable	
  compare	
  to	
  beta	
  sheets.	
  From	
  1M	
  of	
  denaturant	
  concentration,	
  the	
  percentages	
  
of	
  beta	
  sheets	
  decrease	
  to	
  30%	
  while	
  for	
  helixes,	
  even	
  at	
  2M	
  still	
  there	
  are	
  intensities	
  relative	
  

61	
  

to	
  α-­‐helix	
  signal	
  and	
  after	
  4M	
  it	
  drops	
  to	
  about	
  20%	
  of	
  the	
  native	
  (Table	
  II-­‐8),	
  proposing	
  that	
  N-­‐
terminus	
   maybe	
   (partially)	
   fold	
   first.	
   These	
   calculations	
   are	
   all	
   considered	
   with	
   the	
   respect	
   of	
  
220nm	
  regarding	
  beta	
  sheets	
  and	
  222nm/208nm	
  for	
  helixes.	
  	
  
	
  
	
  
	
  

a

c

	
  
	
  
	
  
	
  
	
  
	
  

b

d

	
  
	
  
	
  
	
  
Figure	
   II-­‐22:	
   Equilibrium	
   Unfolding	
   Experiment	
   for	
   wild-­‐type	
   CRBPII	
   in	
   presence	
   of	
  
different	
   concentration	
   of	
   Gd-­‐HCl,	
   a.-­‐b.	
   monitoring	
   by	
   Circular	
   Dichroism	
   (CD).	
   c-­‐d.	
  
Tryptophan	
  fluorescence	
  spectroscopy,	
  right,	
  performed	
  with	
  WT	
  hCRBPII,	
  to	
  study	
  the	
  
presence	
   an	
   intermediate	
   in	
   the	
   folding	
   pathway.	
   	
   Purple	
   solid	
   line	
   represents	
   the	
  
polynomial	
   fit	
  (3rd	
  degree)	
  in	
  b	
  and	
  d.	
  Both	
  b	
   and	
   d	
  spectra	
   are	
  in	
  correlate	
  with	
   previous	
  
studies.	
  (1)	
  	
  	
  
	
  
	
  
	
  

62	
  

Also,	
   based	
   on	
   the	
   previous	
   studies	
   on	
   analysis	
   of	
   three-­‐state	
   protein	
   unfolding	
   data,	
   our	
   CD	
  
spectra	
   of	
   unfolding	
   supports	
   more	
   three-­‐state	
   model	
   than	
   two	
   state.	
   (1)	
   	
   Even,	
   Prof.	
   Lapidus	
  
have	
   tried	
   to	
   fit	
   our	
   data	
   into	
   two-­‐state	
   model	
   and	
   it	
   did	
   not	
   fit	
   correctly.	
   	
   However	
   further	
  
analysis,	
  especially	
  with	
  three-­‐state	
  model	
  fit,	
  need	
  to	
  be	
  done	
  in	
  order	
  to	
  test	
  this	
  hypothesis.	
  
Same	
  set	
  of	
  experiments	
  can	
  be	
  done	
  with	
  the	
  dimer	
  favoring	
  mutants	
  (i.e	
  Y60L),	
  and	
  conduct	
  
them	
   at	
   various	
   monomer-­‐favoring	
   and	
   dimer-­‐favoring	
   concentrations.	
   Observing	
   a	
   similar	
  
intermediate	
  in	
  the	
  folding	
  of	
  these	
  mutants	
  confirms	
  our	
  mechanism	
  even	
  though	
  the	
  folding	
  
product	
   is	
   different	
   and	
   it	
   would	
   allow	
   us	
   manipulating	
   the	
   folding	
   pathway	
   to	
   control	
   the	
  
folding	
  product	
  for	
  this	
  protein,	
  and	
  potentially	
  other	
  members	
  of	
  the	
  iLBP	
  family	
  as	
  well.	
  	
  	
  	
  	
  	
  
Completing	
   mentioned	
   calculations	
   requires	
   more	
   recourses	
   period,	
   however	
   during	
   the	
  
course	
  of	
  my	
  graduate	
  research,	
  were	
  unable	
  to	
  complete	
  the	
  mentioned	
  task.	
  Though	
  we	
  are	
  
optimistic	
  this	
  project	
  will	
  carry	
  on	
  

II-­‐5-­‐3	
  HCRBPII	
  folding	
  route	
  VS	
  other	
  iLBP	
  members	
  
A	
   number	
   of	
   folding	
   studies	
   have	
   been	
   conducted	
   on	
   a	
   few	
   other	
   iLBP	
   family	
   members,	
  
namely	
   hCRBPI	
   and	
   human	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   I	
   (hCRABPI).	
   	
   In	
   the	
   latter	
   case,	
  
a	
  series	
  of	
  experiments	
  have	
  indicated	
  that	
  the	
  folding	
  pathway	
  of	
  hCRABPI	
  involves	
  an	
  early	
  
barrel	
   closure	
   in	
   the	
   folding	
   pathway,	
   even	
   before	
   substantial	
   secondary	
   structure	
   is	
   evident(67-­‐
71)

.	
  	
  
	
  
	
  
	
  
	
  
63	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
Figure	
  II-­‐23:	
  overview	
  picture	
  of	
  hCRABPI	
  folding	
  pathway	
  compare	
  to	
  hCRBPII.	
  	
  
This	
  prevented	
  the	
  protein	
  from	
  forming	
  intermediates	
  that	
  were	
  prone	
  to	
  aggregation	
  
and	
  potentially	
  fibril	
  formation.	
  This	
  mechanism	
  is	
  clearly	
  distinct	
  from	
  what	
  we	
  proposed	
  for	
  
hCRBPII.	
  However,	
  there	
  is	
  no	
  guarantee	
  that	
  all	
  members	
  of	
  a	
  structural	
  family	
  would	
  follow	
  
the	
   same	
   folding	
   pathway.	
   Recently,	
   the	
   crystal	
   structure	
   of	
   a	
   FABP5	
   DS	
   dimer	
   was	
   reported.	
  	
  
The	
  swapped	
  region	
  was	
  almost	
  identical	
  to	
  that	
  of	
  hCRBPII.(93)	
  (Refer	
  to	
  next	
  Chapter,	
  IV)	
  This	
  
shows	
   that	
   other	
   family	
   members	
   ((with	
   only	
   36%	
   sequence	
   identity	
   between	
   FABP5	
   and	
  
hCRBPII)	
  may	
  also	
  have	
  a	
  propensity	
  to	
  form	
  the	
  same	
  DS	
  dimer.	
  The	
  identification	
  of	
  a	
  second,	
  
DS	
   dimer	
   iLBP	
   family	
   member	
   lead	
   us	
   to	
   the	
   idea	
   that	
   a	
   subset	
   of	
   iLBP	
   family	
   members	
   may	
  
indeed	
  be	
  DS	
  dimers	
  in	
  their	
  physiologically	
  relevant	
  forms.	
  

II-­‐6	
  Conformational	
  change	
  driven	
  by	
  ligand	
  binding	
  in	
  hCRBPII	
  
It	
  was	
  previously	
  mentioned	
  that	
  the	
  cellular	
  retinol	
  binding	
  proteins	
  (CRBP’s)	
  are	
  involved	
  in	
  
the	
  trafficking	
  of	
  both	
  retinol	
  and	
  retinal	
  within	
  the	
  cell.	
  Binding	
  studies	
  of	
  hCRBPII	
  DS	
  dimers	
  

64	
  

have	
  demonstrated	
  that	
  DS	
  dimers	
  are	
  competent	
  for	
  ligand	
  binding.(45)	
  (Figure	
   II-­‐9)	
  Knowing	
  
these,	
   in	
   continues	
   collaboration	
   between	
   our	
   lab	
   (prof.	
   Geiger	
   group	
   members)	
   and	
   Prof.	
  
Borhan	
   group;	
   we	
   successfully	
   expressed,	
   purified	
   and	
   determined	
   structures	
   of	
   multiple	
  
variants	
  of	
  holo	
  hCRBPII	
  domain	
  swapped	
  dimer	
  mutants	
  (up	
  to	
  90%	
  dimer	
  formation)	
  bound	
  
Protonated	
  Schiff	
  Base,	
  PSB,	
  to	
  retinal	
  as	
  one	
  of	
  its	
  natural	
  ligand	
  and	
  in	
  the	
  absence	
  of	
  ligand	
  in	
  
that	
   series.	
   (74-­‐76)c	
   	
   (Figure	
   II-­‐	
   25,	
   II-­‐28	
   and	
   II-­‐29)	
   Q108K:T51D	
   holo	
   structure	
   obtained	
   by	
   Dr.	
  
Nosrati	
  and	
  apo	
  form	
  obtained	
  by	
  Alireza	
  Ghanbarpour.	
  (Table	
  II-­‐	
  3)	
  	
  
We	
  found	
  that	
  ligand	
  binding	
  in	
  dimers	
  leads	
  to	
  a	
  reproducible	
  and	
  noticeable	
  conformational	
  
change.	
   Among	
   those,	
   all	
   of	
   the	
   apo	
   structures	
   reveal	
   a	
   symmetric	
   dimer	
   very	
   similar	
   to	
   the	
  
other	
   symmetric	
   dimers	
   previously	
   mentioned	
   in	
   section	
   II-­‐4-­‐2	
   (Figure	
   II-­‐24),	
   though	
   some	
   of	
  
them	
  might	
  differ	
  in	
  their	
  crystal	
  packing.	
  Therefore,	
  the	
  data	
  indicate	
  the	
  symmetric	
  dimer	
  to	
  
be	
   the	
   most	
   common	
   and	
   robust	
   form	
   of	
   DS	
   dimer	
   structure,	
   with	
   the	
   asymmetric	
   dimer	
   a	
  
special	
  case	
  caused	
  by	
  the	
  Y60W	
  mutation.	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
65	
  

Table	
  II-­‐3:	
  	
  Summary	
  of	
  DS	
  dimer	
  crystal	
  structures	
  of	
  all	
  hCRBPII	
  mutants	
  
	
  
	
  
	
  
#	
  molecules	
  
	
  
	
  
Mutants	
  
Holo/Apo	
  
per	
  
Symmetric/	
  
	
  
Q108K:T51D	
  
Q108K:T51D	
  
Q108K:K40D	
  

Apo	
  
Holo	
  
Holo	
  

asymmetric	
  
unit	
  

asymmetric	
  

1	
  
12	
  
12	
  

Symmetric	
  
Asymmetric	
  
Asymmetric	
  

	
  
	
  
	
  

On	
   the	
   other	
   hand,	
   structures	
   of	
   the	
   retinylidene-­‐bound	
   dimers	
   reveal	
   different	
   relative	
  
orientation	
   of	
   the	
   two	
   domains	
   of	
   the	
   dimer	
   in	
   them	
   when	
   compared	
   to	
   the	
   apo	
   symmetric	
  
dimers	
  (Figure	
  II-­‐25).	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   II-­‐24:	
   A	
   comparison	
   of	
   four	
   symmetric	
   dimers,	
   Y60L	
   (orange),	
   WT	
  
(red),	
  E72A	
  (cyan)	
  and	
  Q108K,K40D	
  (green).	
  

Further,	
   most	
   of	
   the	
   holo	
   dimers	
   crystallize	
   with	
   between	
   4	
   and	
   12	
   monomers	
   in	
   the	
  
asymmetric	
   unit,	
   all	
   of	
   which	
   exhibit	
   small	
   differences	
   in	
   relative	
   orientation	
   of	
   the	
   two	
  
domains	
  that	
  make	
  up	
  the	
  DS	
  dimer.	
  This	
  indicates	
  that,	
  in	
  contrast	
  to	
  the	
  apo	
  DS	
  dimer,	
  which	
  
seems	
   to	
   exist	
   as	
   a	
   relatively	
   rigid	
   structure,	
   holo	
   dimers	
   have	
   more	
   flexibility	
   along	
   their	
  

66	
  

strand.	
  However,	
  nearly	
  all	
  dimer	
  molecules	
  of	
  holo	
  structures	
  (i.e.	
  6	
  pairs	
  in	
  Q108K:K40D)	
  are	
  
pretty	
  similar	
  to	
  one	
  another.	
  (Figure	
  II-­‐28a-­‐f)	
  	
  
	
  
	
  
	
  
	
  
	
  

11A#

	
  
	
  
	
  
	
  

Figure	
  II-­‐25:	
  Overlay	
  of	
  holo	
  Q108K:K40D	
  (chains	
  F	
  and	
  A,	
  both	
  shown	
  	
  in	
  orange)	
  and	
  apo	
  
Q108K:T51D	
   (shown	
   in	
   blue)	
   showing	
   the	
   large	
   motion	
   of	
   helix	
   1	
   upon	
   ligand	
   binding	
   (about	
  
15	
  Å).	
  Bound	
  retinal	
  molecules	
  in	
  holo	
  structures	
  are	
  shown	
  as	
  transparent	
  spherical	
  models.	
  
Note,	
   complete	
   dimer	
   in	
   the	
   apo	
   structures	
   was	
   generated	
   by	
   crystallographic	
   two-­‐fold	
  
symmetry	
  operation.	
  	
  
	
  
The	
  major	
  cause	
  of	
  this	
  large	
  relative	
  deviation	
  in	
  two	
  domains	
  is	
  related	
  to	
  the	
  position	
  
Thr59	
  and	
  Tyr60.	
  	
  In	
  all	
  of	
  the	
  mentioned	
  apo	
  symmetric	
  dimers,	
  the	
  Asn59	
  side	
  chain	
  is	
  flipped	
  
into	
  the	
  binding	
  cavity	
  and	
  the	
  residue	
  60	
  is	
  flipped	
  out	
  of	
  the	
  binding	
  cavity	
  (except	
  the	
  WT-­‐
hCRBPII	
  dimer	
  has	
  Y60	
  flipped	
  in	
  form).(94)	
  This	
  is	
  in	
  contrast	
  to	
  every	
  hCRBPII	
  monomer	
  (with	
  
over	
   40	
   structures	
   so	
   far	
   determined),	
   where	
   Asn59	
   is	
   always	
   outside	
   the	
   binding	
   pocket.	
   In	
  
holo	
  DS	
  dimer	
  structures	
  the	
  steric	
  bulk	
  of	
  the	
  ligand	
  forces	
  Asn59	
  to	
  rotate	
  out	
  of	
  the	
  binding	
  
pocket,	
   resulting	
   in	
   the	
   rotation	
   of	
   the	
   entire	
   domain. (Figure	
  II-­‐26a	
  and	
  b),	
   and	
   Tyr60	
   is	
   found	
  
inside	
  the	
  binding	
  pocket.	
  	
  

67	
  

	
   a	
  
	
  

LYS108

b	
  
Re<nal

Glu72

	
  

A"

	
  

Tyr60-holo
Asn59-apo

	
  

Asn59-holo
Asn59-dimer

Tyr60-apo

	
  
	
  

Asn59-monomer

c	
  
d	
  

Re#nal

	
  
	
  

Re#nal
LYS108

LYS108

	
  
	
  

Re#nol

Figure	
   II-­‐26:	
   Ligand	
   binding	
   leads	
   Asn59	
   to	
   flip	
   out	
   of	
   the	
   binding	
   pocket	
   due	
   to	
   steric	
  
effect.	
  a.	
  Overlay	
  of	
  Q108K:T51D-­‐apo	
  (green,	
   obtained	
  by	
  ALireza)	
  and	
  Q108K:K40D-­‐holo	
  
(orange).	
  Allosteric	
   conformational	
  change	
  between	
  apo	
  and	
   holo	
  dimers	
  is	
  caused	
  by	
   the	
  
orientation	
   of	
   the	
   Tyr	
   60	
   and	
   Asn	
   59	
   sidechains.	
   Asn59	
   would	
   sterically	
   clash	
   with	
   the	
  
bound	
   retinylidene,	
   which	
  leads	
  to	
   the	
  flipped-­‐out	
   conformation	
  of	
   Asn59	
   and	
   flipped-­‐in	
  
conformation	
   of	
   Tyr60.	
   b.	
   Overlay	
   of	
   WT	
   apo	
   monomer	
   (pink,	
   PDB	
   code:	
   2RCQ)	
   and	
  WT	
  
apo	
   hCRBPII	
   dimer	
   (green,	
   PDB	
   code:	
   4ZH9).	
   The	
   “flipped	
   in”	
   Asn59	
   conformation	
   is	
  
required	
  to	
  adopt	
  the	
  orientation	
  to	
  form	
  the	
  symmetric	
  DS	
  dimer.	
  c.	
  The	
  similar	
  trajectory	
  
of	
   retinal	
   in	
   monomer	
   and	
   dimer	
   structures	
   of	
   hCRBPII	
   mutants.	
   Q108K:K40L	
   monomer	
  
(purple,	
   PDB	
   code:	
   4RUU)	
   Q108K:K40D	
   dimer	
   (orange)	
   and	
   Q108K:K40L:T51DD	
   dimer	
  
(pink,	
   obtained	
  by	
   Dr.Nosrati)	
  which	
  are	
  all	
  bound	
  to	
   retinal	
   via	
  a	
  covalent	
  bond	
   to	
  Lys108.	
  
d.	
   Overlay	
   of	
   WT	
  monomer	
  hCRBPII	
  bound	
  to	
   retinol	
  (green,	
  PDB	
   code:	
   4QZT)	
   with	
   retinal-­‐
bound	
  Q108K:K40D	
  dimer	
  (orange).	
  
	
  
As	
   shown	
   (Figure	
   II-­‐26c	
   and	
   d),	
   the	
   trajectory	
   of	
   the	
   retinylidene	
   through	
   the	
   dimer	
  
binding	
   pocket	
   is	
   very	
   similar	
   to	
   that	
   seen	
   in	
   both	
   retinol-­‐bound	
   and	
   retinal	
   PSB-­‐bound	
  
monomers,	
  indicating	
  that	
  dimers	
  bind	
  ligands	
  very	
  similar	
  to	
  that	
  of	
  the	
  monomer	
  structures,	
  
whether	
  the	
  ligand	
  is	
  a	
  retinol	
  or	
  retinal.	
  	
  

68	
  

Interestingly,	
  other	
  than	
  the	
  bound	
  retinyldine	
  mutants,	
  the	
  conformational	
  change	
  of	
  
the	
   two	
   domains	
   can	
   be	
   observed	
   for	
   WT-­‐hCRBPII	
   DS	
   dimer	
   bound	
   with	
   retinal	
   (obtained	
   by	
  
Alireza	
   Ghanbarpour)	
   as	
   shown	
   in	
   Figure	
   II-­‐27.	
   Probably	
   due	
   to	
   the	
   low	
   resolution	
   of	
   this	
  
structure	
   (about	
   3.4	
   Å)	
   the	
   electron	
   density	
   of	
   the	
   ligand	
   is	
   weak.	
   	
   However,	
   a	
   strong	
   hydrogen	
  
bond	
  between	
  Tyr60	
  and	
  Glu72	
  	
  is	
  apparently	
  required	
  for	
  the	
  conformational	
  change	
  (similar	
  
to	
  that	
  seen	
  in	
  apo	
  WT-­‐hCRBPII,	
  PDB	
  ID	
  4ZH9)	
  in	
  all	
  holo	
  DS	
  dimers	
  (Figure	
  II-­‐26a	
  and	
  28g-­‐l)	
  To	
  
test	
   this	
   new	
   hypothesis	
   around	
   Tyr60,	
   Q108K:K40L:51F:Y60A	
   mutant	
   structure	
   was	
  
determined	
   by	
   Alireza	
   Ghanbarpour.and	
   showed	
   Thr	
   59	
   position	
   pointing	
   into	
   the	
   binding	
  
pocket.(Data	
  not	
  shown)	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   II-­‐27:	
   Overlay	
   of	
   WT-­‐hCRBPII	
   (holo)	
   DS	
   dimer	
   shown	
   by	
   dark	
   blue	
   and	
  
WT-­‐hCRBPII	
  (apo)	
  (PDB	
  code:4ZH9)	
  shown	
  by	
  green.	
  	
  	
  
This	
  reveals	
  the	
  essential	
  role	
  that	
  Tyr	
  60,	
  and	
  presumably	
  Asn59,	
  play	
  in	
  the	
  mechanism	
  

of	
   allosteric	
   conformational	
   change	
   seen	
   in	
   hCRBPII	
   domain	
   swapped	
   dimers.	
   During	
   my	
  
graduate	
  work,	
  majority	
  of	
  mutants	
  and	
  analysis	
  were	
  done	
  on	
  Tyr60	
  however,	
  	
  more	
  analysis	
  
needs	
  to	
  be	
  carried	
  on	
  Thr59	
  in	
  the	
  future.	
  	
  

69	
  

	
  

Finally,	
   this	
   observation	
   shows	
   DS	
   dimers	
   could	
   potentially	
   	
   be	
   allosterically	
   regulated	
  

during	
   ligand	
   binding,	
   in	
   a	
   way	
   that	
   is	
   not	
   available	
   to	
   iLBP	
   monomeric	
   structures,	
   which	
  
invariably	
   lead	
   to	
   only	
   modest	
   changes	
   in	
   conformation(95).	
   In	
   fact	
   allosteric	
   regulation	
   via	
  
ligand	
  binding	
  has	
  been	
  suggested	
  to	
  be	
  important	
  for	
  the	
  function	
  of	
  a	
  number	
  of	
  iLBP	
  family	
  
members	
  (FABP4	
  and	
  CRABPI	
  and	
  etc).(96)	
  (97,	
  98)	
  	
  
Meanwhile,	
  the	
  identification	
  of	
  several	
  mutational	
  “hot	
  spots”	
  and	
  the	
  apparent	
  ease	
  at	
  which	
  
the	
  dominant	
  folding	
  product	
  can	
  be	
  dramatically	
  altered	
  in	
  hCRBPII	
  have	
  led	
  us	
  to	
  ask	
  whether	
  
the	
   physiologically	
   relevant	
   structure	
   for	
   	
   iLBP	
   family	
   members	
   might	
   actually	
   be	
   the	
   DS	
   dimer.	
  
As	
   the	
   other	
   major	
   goals	
   of	
   this	
   project	
   was	
   to	
   use	
   the	
   information	
   gained	
   regarding	
   the	
  
mechanism	
  and	
  determinants	
  of	
  DS	
  dimer	
  formation	
  to	
  identify	
  other	
  family	
  members	
  of	
  such	
  a	
  
subset.	
   The	
   DS	
   dimer	
   would	
   certainly	
   have	
   distinct	
   and	
   important	
   characteristics	
   in	
   vivo,	
  
including	
  its	
  ability	
  to	
  carry	
  two	
  ligands	
  simultaneously,	
  which	
  will	
  be	
  described	
  in	
  more	
  detail	
  
in	
  the	
  next	
  chapter.	
  
	
  

70	
  

	
  

a	
  

b	
  

	
  

	
  
	
  

d	
  

c	
  

	
  
	
  
	
  

e	
  
	
  

f	
  

	
  
	
  
	
  
	
  
	
  
	
  
Figure	
   II-­‐28:	
  	
  Overlay	
  of	
  different	
  chains	
  in	
  holo	
  Q108K:K40D	
  DS	
  dimer.	
  a.	
  Chains	
  A	
  shown	
  by	
  
orange	
  and	
  I	
  shown	
  by	
  green.	
  	
  b.	
  Chains	
  A	
  shown	
  by	
  orange	
  and	
  L	
  shown	
  by	
  blue.	
  c.	
  Chain	
  A	
  
shown	
   by	
   orange	
   and	
   D	
   shown	
   by	
   pale	
   purple	
   d.	
   Chains	
   A	
   shown	
   by	
   orange	
   and	
   G	
   shown	
   by	
  
purple.	
  	
  e.	
  Overlay	
  of	
  chains	
  A	
  shown	
  by	
  orange	
  and	
  B	
  shown	
  by	
  yellow.	
  f.	
  Overlay	
  of	
  chains	
  A	
  
shown	
   by	
   green	
   and	
   G	
   shown	
   by	
   cyan.	
   g-­‐l:	
   E72…Y60	
   Hydrogen	
   boning	
   motifs	
   in	
   Q108K:K40D	
  
holo	
  DS	
  dimer:	
  g.	
  Chains	
  A	
  and	
  F	
  	
  h.	
  Chains	
  D	
  and	
  H.	
  i.	
  Chain	
  L	
  and	
  J.	
  j.	
  Chains	
  B	
  and	
  E	
  k.	
  Chains	
  K	
  
and	
  I.	
  l.	
  Chains	
  C	
  and	
  G	
  

71	
  

Figure	
  II-­‐28	
  (cont’d)	
  

g	
  

h	
  

	
  
Glu72

Tyr60

Tyr60

Tyr60

Glu72

Asn59

Asn59

Asn59

Glu72

	
  

Glu72

Tyr60

	
  

Asn59

	
  

i	
  

j	
  
Tyr60
Glu72

	
  

Asn59
Asn59

	
  
Tyr60

Tyr60
Glu72

Glu72

Tyr60

Asn59
Asn59

	
  

Glu72

	
  

k	
  

l	
  
Tyr60

Glu72

	
  
Tyr60

Glu72
Glu72

Asn59
Asn59 Tyr60

Tyr60

Asn59
Asn59

Glu72

	
  
	
  
	
  

	
  
	
  
	
  
	
  
	
  

72	
  

	
  
	
  

	
  
	
  
	
  

	
  

	
  
	
  

	
  

Figure	
   II-­‐29:	
   	
   UV	
  data	
  spectrum	
  for	
  different	
  dimer	
  mutants	
  with	
  all-­‐trans	
  retinal	
  was	
  carried	
  
by	
   Dr.	
   Santos.	
   Left,	
   UV	
   spectra	
   represents	
   the	
   kinetically	
   formed	
   product	
   between	
   the	
  
hCRBPII	
   dimer	
  and	
   retinal.	
   Right,	
   this	
  panel	
  depicts	
   the	
  formation	
  of	
  both	
  SB	
  and	
  PSB	
  as	
  a	
  
function	
  of	
  time.	
  	
  

II-­‐7	
  Experimental	
  	
  
II-­‐7-­‐1	
  Material	
  and	
  Method:	
  Site-­‐Directed	
  Mutagenesis	
  
The	
   HCRBPII-­‐pET17b	
   plasmid	
   described	
   was	
   used	
   for	
   mutagenesis	
   following	
   the	
  
QuickChange	
  Site-­‐directed	
  Mutagenesis	
  Kit	
  protocol	
  (Agilent	
  Technologies).	
  PCR	
  conditions	
  for	
  
amplification	
  of	
  mutants	
  are	
  specified	
  below:	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  

73	
  

	
  

Table	
  II-­‐4:	
  PCR	
  protocol	
  for	
  hCRABPII	
  and	
  hCRBPII	
  mutagenesis.	
  	
  

50ÎźL	
  

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
The	
   PCR	
   product	
   (50ΟL)	
   was	
   transformed	
   into	
   DH5alpha	
   cells	
   (50ΟL)	
   and	
   grown	
   on	
  
Luria-­‐Bertani	
  (LB)-­‐agar	
  plates	
  supplemented	
  with	
  Ampicillin	
  (75	
  g/mL)	
  for	
  15h.	
  A	
  single	
  colony	
  
was	
  inoculated	
  in	
  15	
  mL	
  LB	
  medium	
  containing	
  100mg/mL	
  ampicillin	
  and	
  grown	
  at	
  37°C	
  while	
  
shaking,	
   for	
   12	
   hours.	
   DNA	
   purification	
   was	
   performed	
   using	
   a	
   QIAGEN	
   Miniprep	
   DNA	
  
purification	
   kit.	
   The	
   average	
   isolated	
   plasmid	
   DNA	
   concentration	
   was	
   120ng/	
   l	
   in	
   30	
   ΟL	
   in	
  
solution.	
  The	
  DNA	
  sequence	
  was	
  verified	
  by	
  the	
  MSU	
  gene	
  sequencing	
  facility	
  using	
  T7	
  primer.	
  	
  
	
  

74	
  

Primers	
  	
  
Y60W	
  
Forward:	
  5ʼ-­‐	
  CACATTCCGCAACTGGGATGTGGATTTCAC-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  GTGAAATCCACATCCCAGTTGCGGAATGTG-­‐3ʼ	
  
	
  
Y60F	
  
Forward:	
  5ʼ-­‐	
  CACATTCCGCAACTTTGATGTGGATTTCAC	
  -­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  GTGAAATCCACATCAAAGTTGCGGAATGTG-­‐3ʼ	
  
	
  
Y60L	
  
Forward:	
  5’-­‐	
  TTCCGCAACCTTGATGTGGAT-­‐3’	
  
Reverse:	
  5’-­‐	
  ATCCACATCAAGGTTGCGGAA-­‐3’	
  
	
  
E72A	
  
Forward:	
  5’-­‐	
  AGAGTTTGACGCGTACACAAAGAG-­‐	
  3’	
  
Reverse:	
  5’-­‐	
  ACTCCAACAGTGAAATCC-­‐	
  3’	
  
	
  
Y60I	
  
Forward:	
  5’-­‐	
  CATTCCGCAACATTGATGTGGATTTC-­‐3’	
  Reverse:	
  5’-­‐
GAAATCCACATCAATGTTGCGGAATG	
  -­‐3’	
  	
  
	
  
Y60T	
  
Forward:	
  5’-­‐CATTCCGCAACACTGATGTGGATTTC-­‐3’	
  	
  
Reverse:	
  5’-­‐GAAATCCACATCAGTGTTGCGGAATG-­‐3’	
  	
  
	
  
Y60E	
  
Forward:	
  5’-­‐TTCCGCAACGAGGATGTGGATTTC-­‐3’	
  Reverse:	
  5’-­‐GAAATCCACATCCTCGTTGCGGAA	
  -­‐
3’	
  	
  
	
  
Y60D	
  
Forward:	
  5'-­‐	
  CATTCCGCAACGATGATGTGGATTTC	
  3'	
  	
  
Reverse:	
  5'-­‐GAAATCCACATCATCGTTGCGGAATG	
  	
  -­‐3'	
  	
  
	
  
Y60Q	
  
Forward:	
  5'-­‐CATTCCGCAACCAGGATGTGGATTTC-­‐3'	
  	
  
Reverse:	
  5'-­‐GAAATCCACATCCTGGTTGCGGAATG-­‐3'	
  	
  
	
  
Y60A	
  
Forward:	
  5’-­‐CATTCCGCAACGCTGATGTGGATTTC	
  -­‐3’	
  Reverse:	
  5’-­‐
GAAATCCACATCAGCGTTGCGGAATG-­‐3’	
  	
  
	
  

75	
  

T56	
  DELETION	
  	
  	
  
Forward:	
  5’	
  CTTCAAGACAAAAACCACTAGCTTCCGCAACTGGGATGTGGAT	
  -­‐3’	
  	
  
Reverse:	
  5’-­‐	
  ATCCACATCCCAGTTGAGGAAGCTAGTGGTTTTTGTCTTGAA-­‐3’	
  	
  
	
  
T56	
  Deleteion-­‐Y60L	
  
Forward:	
  5’-­‐	
  AAGACAAAAACCACTAGCTTCCGCAACCTTGATGTG-­‐3’	
  
Reverse:	
  5’-­‐CACATCAAGGTTGCGGAAGCTAGTGGTTTTTGTCTT-­‐3’	
  
	
  
T56P	
  
Forward:	
  5’-­‐	
  ACCACTAGCCCATTCCGCAAC-­‐3’	
  	
  
Reverse:	
  5’-­‐	
  GTTGCGGAATGGGCTAGTGGT-­‐3’	
  	
  
	
  
F57G	
  
Forward:	
  5’-­‐	
  ACCACTAGCACAGGCCGCAACCTTGAT-­‐3’	
  Reverse:	
  5’-­‐
ATCAAGGTTGCGGCCTGTGCTAGTGGT-­‐3’	
  	
  
	
  
S55W	
  
Forward:	
  5'-­‐GACAAAAACCACTTGGACATTCCGC-­‐3'	
  Reverse:	
  5'-­‐	
  
GCGGAATGTCCAAGTGGTTTTTGTC-­‐3'	
  	
  

II-­‐7-­‐2	
  Protein	
  Expression	
  and	
  Purification	
  of	
  DS	
  dimer	
  mutants	
  
The	
  protein	
  expression	
  and	
  purification	
  were	
  performed	
  as	
  previously	
  mentioned	
  in	
  II-­‐3-­‐
1	
   with	
   the	
   following	
   modifications:	
   To	
   increase	
   the	
   dimer/monomer	
   ratio,	
   the	
   induction	
  
temperature	
  was	
  increased	
  from	
  room	
  temperature	
  to	
  30oC	
  for	
  overnight	
  growth.	
  In	
  the	
  first	
  
step	
  of	
  purification	
  (Q	
  Sepharose	
  Fast	
  Flow	
  resin	
  (GE	
  Health	
  Sciences))	
  the	
  bound	
  protein	
  was	
  
eluted	
   with	
   10	
   mM	
   Tris,	
   150	
   mM	
   NaCl,	
   pH	
   =	
   8.0,	
   the	
   pure	
   fractions	
   desalted	
   by	
   dialysis	
  
overnight	
  against	
  10	
  mM	
  Tris	
  pH=8.0	
  buffer,	
  and	
  loaded	
  on	
  a	
  second	
  anion	
  exchange	
  column	
  	
  
(15Q,	
  GE	
  Health	
  Sciences,	
  BioLogic	
  DuoFlow	
  system)	
  by	
  a	
  gradient	
  run	
  (Table	
  II-­‐5).	
  The	
  low	
  and	
  
high	
   salt	
   fractions	
   were	
   separately	
   concentrated	
   at	
   4°C	
   to	
   5-­‐8	
   mg/mL	
   and	
   further	
   purified	
   by	
  
size	
   exclusion	
   chromatography	
   (Superdex	
   200	
   16/75	
   column,	
   GE	
   Health	
   Sciences,	
   packed	
   in-­‐

76	
  

house)	
  in	
  a	
  buffer	
  containing	
  10	
  mM	
  Tris,	
  150	
  mM	
  NaCl,	
  pH	
  =	
  8.0.	
  	
  
	
  
Table	
   II-­‐5:	
   Anion	
   Exchange	
   purification	
   protocol	
   for	
   hCRBPII.	
  The	
   buffer	
   used	
   for	
   above	
  
protocol	
   is	
  50	
   mM	
   Tris,	
   pH	
   is	
  adjusted	
   automatically.	
  The	
  proteins	
   elute	
   between	
  4%	
   -­‐8	
   %	
  
2M	
  NaCl.	
  

	
  

II-­‐7-­‐3	
  Protein	
  Refolding	
  
The	
  monomeric	
  protein	
  fraction	
  was	
  denatured	
  by	
  addition	
  of	
  urea	
  to	
  8	
  M	
  at	
  4	
  °C	
  and	
  
refolded	
  by	
  dialysis	
  against	
  10	
  mM	
  Tris	
  pH	
  =	
  8.0	
  with	
  a	
  Slide-­‐A-­‐Lyzer®	
  Dialysis	
  Cassette	
  (3,500	
  
MWCO	
  cut	
  off,	
  Thermo	
  Fisher	
  Sci.).	
  The	
  renatured	
  protein	
  was	
  concentrated	
  and	
  analyzed	
  by	
  
size	
  exclusion	
  chromatography.	
  (45)	
  (Table	
  II-­‐1)	
  
	
  
	
  

77	
  

II-­‐7-­‐4	
  Thermal	
  Melting	
  Curves	
  for	
  monomer	
  and	
  dimer	
  mutants.	
  
10	
   ¾M	
   protein	
   (in	
   50	
   mM	
   phosphate	
   buffer,	
   pH	
   =	
   7.5)	
   was	
   heated	
   and	
   monitored	
   by	
   CD	
  
spectroscopy	
   using	
   a	
   JASCO	
   (J-­‐810)	
   Spectropolarimeter	
   connected	
   to	
   a	
   NESLAB	
   RTE-­‐111	
   bath	
  
circulator.	
   The	
   temperature	
   was	
   increased	
   by	
   5°C	
   for	
   every	
   measurement	
   and	
   the	
   CD	
   was	
  
measured.	
  The	
  curves	
  were	
  fitted	
  based	
  on	
  the	
  Levenberg-­‐Marquardt	
  formula	
  using	
  OriginLab	
  
(OriginLab,	
  Northampton,	
  MA,	
  Figure	
  II-­‐6	
  and	
  31)	
  (99,	
  100)	
  
Final	
  concentration	
  of	
  protein	
  for	
  folding	
  studies	
  was	
  15	
  ¾M	
  in	
  folding	
  studies	
  experiments	
  to	
  
see	
   a	
   resalable	
   signal	
   between	
   220-­‐230nm	
   regarding	
   beta	
   sheets	
   however	
   for	
   fluorescence	
  
studies	
   were	
   at	
   9¾M.	
   (Section	
   II-­‐5-­‐2)	
   Excitation	
   wavelength	
   was	
   at	
   280nm	
   and	
   emission	
  
wavelength	
   between	
   320-­‐450nm.	
   (Figure	
  II-­‐22)	
   All	
   of	
   those	
   mentioned	
   experiments	
   were	
   done	
  
in	
  Prof.	
  Borhan	
  and	
  Prof.	
  Lapidus	
  (physics	
  department)	
  labs.	
  	
  
	
  

78	
  

	
  

Monomer	
  KLY60W	
  

aA

Monomer KL-Y60L

b

20

	
  
	
  
	
  

CD (mdeg)

0
Temperature:

-20
-40
-60

	
  

-80
200

	
  

220

240

260

35

75

45

80

55

85

65

90

70

95

280

300

320

Wavelength (nm)

	
  
	
  

Monomer Y60L

c

d 20

20

	
  

Dimer Y60L

10
0

	
  

	
  
	
  
	
  
	
  
	
  

-20

35

65

-30

40

70

-40

50

80

55

-60
200

220

240

260

280

300

-20

Temperature:

-40
-60
-80

90

-100
200

320

220

240

Monomer WT

e

f

20

CD (mdeg)

0

280

50

30

55

35

60

40

80

45

90

300

320

Dimer WT
0

-20

Temperature:

25
35
45
55
65
70

-40
-60
-80
-100
200

260

25

Wavelength (nm)

Wavelength (nm)

	
  
	
  

60

-50

	
  
	
  

30

CD (mdeg)

	
  

Temperature:

-10

CD (mdeg)

	
  

CD (mdeg)

0

75
80
85
90
95

-50
Temperature:

25
30
35
40
45
50
55
60

-100
-150

65
70
75
80
85
90
95

-200

220

240

260

280

300

200

320

220

240

260

280

300

320

Wavelength (nm)

Wavelength (nm)

Figure	
  II-­‐31:	
  Circular	
  Dichroism	
  spectra	
  of	
  various	
  hCRBPII	
  mutants	
  as	
  a	
  function	
  
of	
   temperature.	
   a.	
   Monomer	
   KLY60W.	
   b.	
   Dimer	
   KLY60W.	
   c.	
   Monomer	
   Y60L.	
   d.	
  
Dimer	
   Y60L.	
  e.	
  Monomer	
  WT	
   hCRBPII.	
  f.	
   Dimer	
   WT	
  hCRBPII.	
   	
  Note,	
  most	
  of	
  the	
  
secondary	
   structure	
   is	
   preserved	
   in	
   both	
   the	
   monomers,	
   even	
   at	
   the	
   highest	
  
temperature	
  achieved.	
  	
  Thus	
  melting	
  curves	
  can	
  only	
  give	
  a	
  lower	
  limit	
  for	
  the	
  TM,	
  
consistent	
  with	
  the	
  view	
  that	
  the	
  monomers	
  are	
  more	
  stable	
  than	
  the	
  dimers.	
  	
  
	
  
79	
  

Table	
  II-­‐6:	
  Calculated	
  percentage	
  of	
  helix/strand	
  for	
  each	
  denaturant	
  (GD.HCl)	
  concentration	
  
during	
  unfolding	
  experiment.	
  220nm	
  corresponds	
  to	
  beta	
  sheets.	
  222nm/208nm	
  corresponds	
  
to	
  α-­‐helixes	
  
	
  
Concentration	
  
of	
  denaturant	
  

	
  

0M	
  	
  
0.6	
  
0.8	
  
1.0	
  
1.2	
  
1.4	
  
2.0	
  
2.8	
  
4.0	
  
6	
  

	
  
	
  
	
  
	
  
	
  

mdeg	
  
at	
  
220nm	
  
-­‐20	
  	
  
-­‐17	
  
-­‐14	
  
-­‐7	
  
-­‐6	
  
-­‐4	
  
-­‐4	
  
-­‐5	
  
-­‐4	
  
-­‐2	
  

%	
  of	
  	
  
mdeg	
  at	
  
220nm	
  
100	
  
85	
  
70	
  
35	
  
30	
  
20	
  
20	
  
25	
  
20	
  
10	
  

mdeg	
  at	
  
222nm/
208nm	
  
-­‐12/-­‐14	
  	
  
-­‐11/-­‐11	
  	
  	
  	
  
-­‐8/-­‐9	
  
-­‐6/-­‐10	
  	
  	
  	
  
-­‐5/-­‐12	
  
-­‐3/-­‐9	
  
-­‐3/-­‐11	
  	
  	
  
-­‐4/-­‐14	
  	
  	
  	
  
-­‐3/-­‐7	
  	
  	
  
-­‐2/-­‐4	
  	
  	
  	
  

%	
  of	
  mdeg	
  at	
  
222nm/208nm	
  
100%	
  
100/78	
  
66/64	
  
50/71	
  
41/85	
  
25/64	
  
25/78	
  
30/100	
  
25/50	
  
16/26	
  

II-­‐7-­‐5	
  	
  Kd	
  determination	
  
The	
   dissociation	
   constant	
   (Kd)	
   of	
   both	
   monomer	
   and	
   dimer	
   species	
   with	
   all-­‐trans-­‐retinal	
  
and	
   all-­‐trans-­‐retinol	
   were	
   determined	
   by	
   fluorescence	
   titration	
   as	
   previously	
   explained	
   in	
  
chapter	
  I.	
  	
  

II-­‐7-­‐6	
  Extinction	
  Coefficient	
  Determination	
  	
  
The	
  absorption	
  extinction	
  coefficients	
  for	
  the	
  various	
  CRBPII	
  mutants	
  were	
  determined	
  
according	
  to	
  the	
  method	
  described	
  by	
  Gill	
  and	
  von	
  Hippel	
  (101).	
  	
  

II-­‐7-­‐7	
  Crystallization,	
  data	
  collection	
  and	
  refinement	
  
Each	
  mutant	
  protein	
  was	
  concentrated	
  to	
  between	
  4	
  to	
  6	
  mg/mL	
  in	
  a	
  buffer	
  containing	
  
10	
   mM	
   Tris,	
   100	
   mM	
   NaCl,	
   pH	
   =	
   8.0.	
   Crystals	
   were	
   grown	
   using	
   the	
   hanging	
   drop	
   vapor	
  
diffusion	
  method	
  using	
  1	
  ΟL	
  of	
  protein	
  solution	
  and	
  1	
  ΟL	
  of	
  crystallization	
  solution	
  in	
  the	
  drop	
  
and	
  1	
  mL	
  of	
  crystallization	
  solution	
  in	
  the	
  reservoir.	
  All	
  of	
  the	
  crystallization	
  trials	
  with	
  retinal	
  

80	
  

were	
  wrapped	
  in	
  aluminum	
  foil	
  to	
  protect	
  the	
  light	
  sensitive	
  chromophore.	
  	
  The	
  best	
  crystals	
  
grew	
   using	
   a	
   crystallization	
   solution	
   of	
   25-­‐30%	
   PEG4000,	
   0.1	
   M	
   NaCH3CO2Na	
   pH	
   4.2,	
   0.1	
   M	
  
NH4CH3CO2.	
  The	
  crystals	
  were	
  soaked	
  in	
  a	
  cryoprotectant	
  solution	
  (30%	
  PEG4000,	
  0.1	
  M	
  sodium	
  
acetate	
  pH	
  =	
  4.0,	
  0.1	
  M	
  ammonium	
  acetate	
  with	
  15%	
  glycerol),	
  flash	
  frozen	
  in	
  loops	
  in	
  liquid	
  
nitrogen	
  and	
  stored	
  in	
  a	
  liquid	
  nitrogen	
  Dewar	
  prior	
  to	
  data	
  collection.	
  
Diffraction	
   data	
   were	
   collected	
   at	
   the	
   Advanced	
   Photon	
   Source	
   (APS)	
   (Argonne	
   National	
  
Laboratory	
   IL)	
   LS-­‐CAT,	
   (sector	
   21-­‐ID-­‐D,F,G)	
   using	
   a	
   MAR300	
   detector	
   and	
   1.00Å	
   wavelength	
  
radiation	
   at	
   100K.	
   The	
   initial	
   diffraction	
   data	
   were	
   indexed,	
   processed	
   and	
   scaled	
   using	
   the	
  
HKL2000	
   software	
   package.(32)	
   The	
   structures	
   were	
   solved	
   by	
   molecular	
   replacement	
   using	
  
MOLREP	
  in	
  CCP4	
  or	
  PHASER	
  in	
  PHENIX	
  and	
  hCRBPII	
  (PDB	
  entry	
  2RCQ)	
  as	
  a	
  search	
  model.(22)	
  The	
  
initial	
  electron	
  density	
  map	
  was	
  produced	
  by	
  REFMAC5	
  in	
  the	
  CCP4	
  package	
  or	
  Phaser-­‐MR	
  in	
  
PHENIX.(33,	
   102,	
   103)	
   Model	
   rebuilding,	
   placement	
   of	
   water	
   molecules	
   etc.	
   were	
   done	
   using	
  
COOT.(36)	
   	
   The	
   structures	
   were	
   refined	
   using	
   either	
   the	
   CCP4	
   and	
   PHENIX	
   program	
   packages.	
  
The	
  chromophore	
  was	
  built	
  into	
  the	
  KLY60W	
  mutant	
  structures	
  using	
  the	
  JLIGAND	
  program	
  to	
  
generate	
   restraints.(104,	
   105)	
   Crystallographic,	
   atomic	
   coordinates	
   and	
   structure	
   factors	
   have	
  
been	
  deposited	
  in	
  the	
  Protein	
  Data	
  Bank,	
  www.pdb.org.	
  
	
  
	
  
	
  
	
  
	
  
	
  

81	
  

	
  
Table	
  II-­‐7:	
  X-­‐ray	
  crystallographic	
  data	
  and	
  refinement	
  statistics	
  for	
  monomeric	
  
and	
  dimeric	
  hCRBPII	
  mutants.	
  
	
  

	
  

Q108K:K40L:Y60W	
  
(monomer)	
  

Y60L	
  (monomer)	
  

P1	
  
30.08	
  
36.05	
  

P1	
  
30.27	
  
36.38	
  

c	
  (Å)	
  

63.92	
  
90.94	
  
91.99	
  
113.93	
  
2	
  

126.10	
  
85.43	
  
85.18	
  
66.65	
  
4	
  

124534	
  
41751	
  
	
   a
92.8(68.0) 	
  
	
  a
50.30(3.58) 	
  
a
4.1(29.3) 	
  

268817	
  
77790	
  
a
94.7(94.6) 	
  
a
43.37(1.64) 	
  
a
3.3(77.6) 	
  

Average	
  I/σ	
  
Rmerge(%)	
  
Resolution	
  (Å)	
  (Last	
  Shell)	
  
Rwork/	
  Rfree	
  (%)	
  
RMSD	
  from	
  ideal	
  value	
  

a

21.30-­‐1.47(1.52-­‐1.50) 	
  
19.86/25.06	
  
	
  

Bond	
  Length	
  (Å)	
  
0.025	
  
Bond	
  Angle	
  (°)	
  
2.36	
  
Average	
  B	
  factor	
  
19.98	
  
Number	
  of	
  water	
  
228	
  
molecules	
  
PDB	
  ID	
  
4ZJ0	
  
a	
  
Values	
  in	
  the	
  parenthesis	
  refer	
  to	
  the	
  last	
  resolution	
  shell.	
  

	
  
	
  

Space	
  group	
  
a(Å)	
  
b(Å)	
  
ι(°	
  )	
  
β(°	
  )	
  
δ(°	
  )	
  
Molecules	
  per	
  
Asymmetric	
  Unit	
  
Total	
  reflection	
  
Unique	
  Reflection	
  
Completeness	
  (%)	
  

	
  

	
  
	
  
	
  
	
  
	
  
	
  
a

50-­‐1.50(1.53-­‐1.50) 	
  
22.72/24.74	
  

	
  
	
  

0.0109	
  
1.22	
  
30.86	
  
459	
  

	
  

5DG4	
  

	
  

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

82	
  

Table	
  II-­‐7	
  (cont’d)	
  
!

Q108K:!
K40L:Y6
0W"
(dimer)!
P21212!
62.77!
109.22!

Y60W"
(dimer)!

Y60L"
(dime
r)!

E72A"
(dim
er)!

WT#!
(dim
er)!

P21212!
62.62!
109.77!

36.39!
90.00!
90.00!
90.00!
2!

36.32!
90.00!
90.00!
90.00!
2!

P2121
63.60
2!
3!
60.42
4!
36.77
6!
90.00!
90.00!
90.00!
1!

P21
262.
1 2!
238!
72.
901!
60.
487!
90.
00!
90.
00!
90.
00!
2!

P212
64.0
1 2!
21!
60.5
44!
36.8
51!
90.0
0!
90.0
0!
90.0
0!
1!

288797!
20048!
!
99.9(10
a
0.0) !
78.16(6
a
4.7(46.
.23) !
a
5) !

345311!
55151!
98.6(10
!a
0.0) !
66.13(7
!a
5.2(31.
.71) !
!a
7) !

30108
21293!
5!
100(9
a
9.81)
82.03 !
(3.18)
4.1(6
a a
9.5)
! !

3807
2724
39!
2!
100(
a
100)
75.1 !
3(2.7
5.5(8
4.7)
6)a!a!

3389
4455!
1!
100(
99.7
34.6
a
0) !
3(2.4
5.9(7
3.1)
1)a!a!

Resolution*(Å)*
(Last&Shell)!

50.00"
1.91(1.
94"
a
1.91) !

50.00"
1.70(1.
73"
a
1.70) !

50.00"
1.55(1.58
a
"1.55) !

50"
1.77(1.8
a
"1.77) !

Rwork/"Rfree$(%)!

18.00/2
1.84!
!
!

18.17/2
3.94!
!
!
!

19.91
/24.5
7!

17.9
4/22.
41!

	
  
Space&group!
a(Å)!
b(Å)!
c"(Å)!
ι(°#)!
β(°#)!
δ(°#)!
Molecules(per!
Asymmetric*
Unit!
Total&reflection!
Unique'
Reflection!
Completeness*
(%)!
Average'I/σ!
Rmerge(%)!

	
  

	
  
!
	
  !
!
!

	
   !

!

!

!

!

!

!

!

	
  
	
  
	
  
	
  
	
  

83	
  

	
  

P212121!
65.47!
73.60!

	
  

351.60! 	
  
90.00!
90.00! 	
  
90.00!
12!
	
  
2363956! 	
  
53735!
99.80%
	
  
a
(100.0) !
26.2(2.7
a
5) !
3.5(30.0)
a
	
  
!

50.0
0"
2.66(
2.71"
2.66)
a
!
17.5
7/26.
28!

19.92"
2.60(2.64"
a
2.60) !

!

!
!

	
  

Q108K:!
K40D!
(dimer)!

17.66/28
.05!
!
!

	
  

	
  

Table	
  II-­‐7	
  (cont’d)	
  
RMSD%from%ideal%
values!
Bond%Length%(Å)!

!

!

!

0.008!

0.007!

0.013!

0.002!

0.008!

!Bond%Angle%(°)!

1.11!

1.03!

1.28!

0.51!

1.24!

Average'B'factor!

28.71!

34.31!

34.17!

36.71!

55.85! 	
  

Number'of'water!
molecules!

352!

0
.
0
2
3!
2
.
1
1!
1
9
.
9
2!
2
1
7!

147!

462!

50!

PDB$ID!

4ZR2!

4
Z
C
B!

4ZH6!

5DPQ!

	
  

	
  

	
  

	
  
	
  

!

!
4ZH9!!!!!!!!5U6G!
!

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

84	
  

241! 	
  
!
!

	
  

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

REFERENCES	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

85	
  

1.	
  
	
  
2.	
  
	
  
3.	
  
	
  
4.	
  
	
  
5.	
  
	
  
6.	
  
	
  
7.	
  
	
  
8.	
  
	
  
9.	
  
	
  
10.	
  
	
  
11.	
  
	
  
12.	
  

REFERENCES	
  
	
  
	
  
	
  
Bennett,	
   M.	
   J.,	
   Schlunegger,	
   M.	
   P.,	
   and	
   Eisenberg,	
   D.	
   (1995)	
   3D	
   domain	
   swapping	
   -­‐	
   A	
  
mechanism	
  for	
  oligomer	
  assembly,	
  Protein	
  Science	
  4,	
  2455-­‐2468.	
  
Crestfield,	
   A.	
   M.,	
   Stein,	
   W.	
   H.,	
   and	
   Moore,	
   S.	
   (1962)	
   On	
   the	
   aggregation	
   of	
   bovine	
  
pancreatic	
  ribonuclease,	
  Archives	
  of	
  biochemistry	
  and	
  biophysics	
  Suppl	
  1,	
  217-­‐222.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   As	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  Science	
  11,	
  1285-­‐1299.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   Structures	
   of	
   the	
   two	
   3D	
  
domain-­‐swapped	
  RNase	
  A	
  trimers,	
  Protein	
  Science	
  11,	
  371-­‐380.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   S.	
   (2001)	
   Crystal	
   structure	
   of	
   an	
  
Rnase	
  A	
  dimer	
  displaying	
  a	
  new	
  type	
  of	
  3D	
  domain	
  swapping,	
  Nucleic	
  Acid	
  Database.	
  
Liu,	
  Y.	
  S.,	
  Hart,	
  P.	
  J.,	
  Schlunegger,	
  M.	
  P.,	
  and	
  Eisenberg,	
  D.	
  (1998)	
  The	
  crystal	
  structure	
  of	
  
a	
   3D	
   domain-­‐swapped	
   dimer	
   of	
   RNase	
   A	
   at	
   a	
   2.1-­‐angstrom	
   resolution,	
   Proceedings	
   of	
  
the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  95,	
  3437-­‐3442.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   as	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  11,	
  1285-­‐1299.	
  
Koharudin,	
   L.	
   M.,	
   Liu,	
   L.,	
   and	
   Gronenborn,	
   A.	
   M.	
   (2013)	
   Different	
   3D	
   domain-­‐swapped	
  
oligomeric	
   cyanovirin-­‐N	
   structures	
   suggest	
   trapped	
   folding	
   intermediates,	
   Proc	
   Natl	
  
Acad	
  Sci	
  U	
  S	
  A	
  110,	
  7702-­‐7707.	
  
Assar,	
  Z.,	
  Nossoni,	
  Z.,	
  Wang,	
  W.,	
  Santos,	
  E.	
  M.,	
  Kramer,	
  K.,	
  McCornack,	
  C.,	
  Vasileiou,	
  C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   Domain-­‐Swapped	
   Dimers	
   of	
   Intracellular	
   Lipid-­‐
Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure	
  24,	
  1590-­‐1598.	
  
Bax,	
  B.,	
  Lapatto,	
  R.,	
  Nalini,	
  V.,	
  Driessen,	
  H.,	
  Lindley,	
  P.	
  F.,	
  Mahadevan,	
  D.,	
  Blundell,	
  T.	
  L.,	
  
and	
   Slingsby,	
   C.	
   (1990)	
   X-­‐ray	
   analysis	
   of	
   beta-­‐B2-­‐crystallin	
   and	
   evolution	
   of	
   oligomeric	
  
lens	
  proteins,	
  Nature	
  347,	
  776-­‐780.	
  
Hakansson,	
  M.,	
  Svensson,	
  A.,	
  Fast,	
  J.,	
  and	
  Linse,	
  S.	
  (2001)	
  An	
  extended	
  hydrophobic	
  core	
  
induces	
  EF-­‐hand	
  swapping,	
  Protein	
  Science	
  10,	
  927-­‐933.	
  
Rousseau,	
   F.,	
   Schymkowitz,	
   J.,	
   and	
   Itzhaki,	
   L.	
   S.	
   (2012)	
   Implications	
   of	
   3D	
   domain	
  
swapping	
   for	
   protein	
   folding,	
   misfolding	
   and	
   function,	
   In	
   Protein	
   Dimerization	
   and	
  
Oligomerization	
  in	
  Biology	
  (Mattews,	
  J.	
  M.,	
  Ed.),	
  pp	
  137-­‐152.	
  
86	
  

13.	
  

	
  
14.	
  

	
  
15.	
  
	
  
16.	
  
	
  
16.	
  
	
  
17.	
  
	
  
18.	
  
	
  
19.	
  
	
  
20.	
  
	
  
21.	
  
	
  
22.	
  

Yang,	
   F.,	
   Bewley,	
   C.	
   A.,	
   Louis,	
   J.	
   M.,	
   Gustafson,	
   K.	
   R.,	
   Boyd,	
   M.	
   R.,	
   Gronenborn,	
   A.	
   M.,	
  
Clore,	
  G.	
  M.,	
  and	
  Wlodawer,	
  A.	
  (1999)	
  Crystal	
  structure	
  of	
  cyanovirin-­‐N,	
  a	
  potent	
  HIV-­‐
inactivating	
  protein,	
  shows	
  unexpected	
  domain	
  swapping,	
  Journal	
  of	
  Molecular	
  Biology	
  
288,	
  403-­‐412.	
  
Szilagyi,	
   A.,	
   Zhang,	
   Y.,	
   and	
   Zavodszky,	
   P.	
   (2012)	
   Intra-­‐Chain	
   3D	
   Segment	
   Swapping	
  
Spawns	
  the	
  Evolution	
  of	
  New	
  Multidomain	
  Protein	
  Architectures,	
  Journal	
  of	
  Molecular	
  
Biology	
  14.	
   Szilagyi,	
   A.,	
   Zhang,	
   Y.,	
   and	
   Zavodszky,	
   P.	
   (2012)	
   Intra-­‐Chain	
   3D	
   Segment	
  
Swapping	
   Spawns	
   the	
   Evolution	
   of	
   New	
   Multidomain	
   Protein	
   Architectures,	
   Journal	
   of	
  
Molecular	
  Biology	
  415,	
  221-­‐235.	
  
Stroud,	
   J.	
   C.,	
   Liu,	
   C.,	
   Teng,	
   P.	
   K.,	
   and	
   Eisenberg,	
   D.	
   (2012)	
   Toxic	
   fibrillar	
   oligomers	
   of	
  
amyloid-­‐beta	
   have	
   cross-­‐beta	
   structure,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  109,	
  7717-­‐7722.	
  
Cleary,	
   J.	
   P.,	
   Walsh,	
   D.	
   M.,	
   Hofmeister,	
   J.	
   J.,	
   Shankar,	
   G.	
   M.,	
   Kuskowski,	
   M.	
   A.,	
   Selkoe,	
   D.	
  
J.,	
   and	
   Ashe,	
   K.	
   H.	
   (2005)	
   Natural	
   oligomers	
   of	
   the	
   amyloid-­‐protein	
   specifically	
   disrupt	
  
cognitive	
  function,	
  Nature	
  Neuroscience	
  8,	
  79-­‐84.	
  
Stroud,	
   J.	
   C.,	
   Liu,	
   C.,	
   Teng,	
   P.	
   K.,	
   and	
   Eisenberg,	
   D.	
   (2012)	
   Toxic	
   fibrillar	
   oligomers	
   of	
  
amyloid-­‐beta	
   have	
   cross-­‐beta	
   structure,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  109,	
  7717-­‐7722.	
  
Cleary,	
   J.	
   P.,	
   Walsh,	
   D.	
   M.,	
   Hofmeister,	
   J.	
   J.,	
   Shankar,	
   G.	
   M.,	
   Kuskowski,	
   M.	
   A.,	
   Selkoe,	
   D.	
  
J.,	
   and	
   Ashe,	
   K.	
   H.	
   (2005)	
   Natural	
   oligomers	
   of	
   the	
   amyloid-­‐protein	
   specifically	
   disrupt	
  
cognitive	
  function,	
  Nature	
  Neuroscience	
  8,	
  79-­‐84.	
  
Fandrich,	
   M.,	
   Fletcher,	
   M.	
   A.,	
   and	
   Dobson,	
   C.	
   M.	
   (2001)	
   Amyloid	
   fibrils	
   from	
   muscle	
  
myoglobin	
   -­‐	
   Even	
   an	
   ordinary	
   globular	
   protein	
   can	
   assume	
   a	
   rogue	
   guise	
   if	
   conditions	
  
are	
  right,	
  Nature	
  410,	
  165-­‐166.	
  
Gronenborn,	
   A.	
   M.	
   (2009)	
   Protein	
   acrobatics	
   in	
   pairs	
   -­‐	
   dimerization	
   via	
   domain	
  
swapping,	
  Current	
  Opinion	
  in	
  Structural	
  Biology	
  19,	
  39-­‐49.	
  
Schlunegger,	
  M.	
  P.,	
  Bennett,	
  M.	
  J.,	
  and	
  Eisenberg,	
  D.	
  (1997)	
  Oligomer	
  formation	
  by	
  3D	
  
domain	
   swapping:	
   A	
   model	
   for	
   protein	
   assembly	
   and	
   misassembly,	
   In	
   Advances	
   in	
  
Protein	
  Chemistry,	
  Vol	
  50:	
  Protein	
  Misassembly	
  (Wetzel,	
  R.,	
  Ed.),	
  pp	
  61-­‐122.	
  
Selkoe,	
  D.	
  J.	
  (2004)	
  Folding	
  proteins	
  in	
  fatal	
  ways	
  (vol	
  426,	
  pg	
  900,	
  2003),	
  Nature	
  428,	
  
445-­‐445.	
  
Shtilerman,	
  M.	
  D.,	
  Ding,	
  T.	
  T.,	
  and	
  Lansbury,	
  P.	
  T.	
  (2002)	
  Molecular	
  crowding	
  accelerates	
  
fibrillization	
   of	
   alpha-­‐synuclein:	
   Could	
   an	
   increase	
   in	
   the	
   cytoplasmic	
   protein	
  
concentration	
  induce	
  Parkinson's	
  disease?,	
  Biochemistry	
  41,	
  3855-­‐3860.	
  
87	
  

23.	
  
	
  
24.	
  
	
  
25.	
  
	
  
26.	
  
	
  
27.	
  
	
  
28.	
  
	
  
29.	
  
	
  
30.	
  
31.	
  
	
  
32.	
  
	
  
33.	
  
	
  
34.	
  
	
  
35.	
  

Sunde,	
  M.,	
  and	
  Blake,	
  C.	
  (1997)	
  The	
  structure	
  of	
  amyloid	
  fibrils	
  by	
  electron	
  microscopy	
  
and	
   X-­‐ray	
   diffraction,	
   In	
   Advances	
   in	
   Protein	
   Chemistry,	
   Vol	
   50:	
   Protein	
   Misassembly	
  
(Wetzel,	
  R.,	
  Ed.),	
  pp	
  123-­‐159.	
  
Bennett,	
  M.	
  J.,	
  Choe,	
  S.,	
  and	
  Eisenberg,	
  D.	
  (1994)	
  Refined	
  structure	
  of	
  dimeric	
  diphtheria	
  
toxin	
  at	
  2.0	
  A	
  resolution,	
  Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  3,	
  1444-­‐
1463.	
  
O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
  environment,	
  Curr	
  Opin	
  Struct	
  Biol	
  11,	
  114-­‐119.	
  
Bennett,	
   M.	
   J.,	
   Choe,	
   S.,	
   and	
   Eisenberg,	
   D.	
   (1994)	
   Rlefined	
   strucutre	
   of	
   dimeric	
  
diphtheria-­‐toxin	
  at	
  2.0-­‐angstrom	
  resolution,	
  Protein	
  Science	
  3,	
  1444-­‐1463.	
  
Carroll,	
   S.	
   F.,	
   Barbieri,	
   J.	
   T.,	
   and	
   Collier,	
   R.	
   J.	
   (1986)	
   Dimeric	
   form	
   of	
   diphtheria-­‐toxin-­‐
purification	
  and	
  characterization,	
  Biochemistry	
  25,	
  2425-­‐2430.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
   environment	
   (vol	
   11,	
   pg	
   114,	
   2001),	
   Current	
   Opinion	
   in	
   Structural	
   Biology	
  
11,	
  500-­‐500.	
  
Minton,	
   A.	
   P.	
   (2000)	
   Implications	
   of	
   macromolecular	
   crowding	
   for	
   protein	
   assembly,	
  
Current	
  Opinion	
  in	
  Structural	
  Biology	
  10,	
  34-­‐39.	
  
O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   J.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nature	
  chemical	
  biology	
  10,	
  884-­‐891.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2004)	
  Monitoring	
  protein	
  stability	
  and	
  aggregation	
  in	
  
vivo	
  by	
  real-­‐time	
  fluorescent	
  labeling,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  101,	
  523-­‐528.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2006)	
  Inhibition	
  of	
  protein	
  aggregation	
  in	
  vitro	
  and	
  in	
  
vivo	
  by	
  a	
  natural	
  osmoprotectant,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  103,	
  13357-­‐13361.	
  
Ropson,	
  I.	
  J.,	
  Boyer,	
  J.	
  A.,	
  Schaeffer,	
  B.	
  A.,	
  and	
  Dalessio,	
  P.	
  M.	
  (2009)	
  Comparison	
  of	
  the	
  
folding	
   mechanism	
   of	
   highly	
   homologous	
   proteins	
   in	
   the	
   lipid-­‐binding	
   protein	
   family,	
  
Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  75,	
  799-­‐806.	
  

	
  
88	
  

36.	
  
	
  
37.	
  
	
  
38.	
  
	
  
39.	
  
	
  
40.	
  

	
  
41.	
  
	
  
42.	
  
	
  
43.	
  
	
  
44.	
  
	
  
45.	
  
	
  
46.	
  
	
  
47.	
  

Burns,	
   L.	
   L.,	
   and	
   Ropson,	
   I.	
   J.	
   (2001)	
   Folding	
   of	
   intracellular	
   retinol	
   and	
   retinoic	
   acid	
  
binding	
  proteins,	
  Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  43,	
  292-­‐302.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nat.	
  Chem.	
  Biol.	
  10,	
  884-­‐891.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins-­‐-­‐insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  lipid	
  research	
  43,	
  328-­‐349.	
  
Lucke,	
   C.,	
   Huang,	
   S.,	
   Rademacher,	
   M.,	
   and	
   Ruterjans,	
   H.	
   (2002)	
   New	
   insights	
   into	
  
intracellular	
   lipid	
   binding	
   proteins:	
   The	
   role	
   of	
   buried	
   water,	
   Protein	
   science	
   :	
   a	
  
publication	
  of	
  the	
  Protein	
  Society	
  11,	
  2382-­‐2392.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  crystallographica.	
  Section	
  
D,	
  Biological	
  crystallography	
  70,	
  3226-­‐3232.	
  
Schaap,	
   F.	
   G.,	
   van	
   der	
   Vusse,	
   G.	
   J.,	
   and	
   Glatz,	
   J.	
   F.	
   (2002)	
   Evolution	
   of	
   the	
   family	
   of	
  
intracellular	
   lipid	
   binding	
   proteins	
   in	
   vertebrates,	
   Molecular	
   and	
   cellular	
   biochemistry	
  
239,	
  69-­‐77.	
  
Smathers,	
   R.	
   L.,	
   and	
   Petersen,	
   D.	
   R.	
   (2011)	
   The	
   human	
   fatty	
   acid-­‐binding	
   protein	
   family:	
  
evolutionary	
  divergences	
  and	
  functions,	
  Human	
  genomics	
  5,	
  170-­‐191.	
  
Crist,	
   R.	
   M.,	
   Vasileiou,	
   C.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2006)	
  
Engineering	
  a	
  Rhodopsin	
  Protein	
  Mimic,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  128,	
  
4522-­‐4523.	
  
Vasileiou,	
   C.,	
   Vaezeslami,	
   S.,	
   Crist,	
   R.	
   M.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2007)	
   Protein	
   design:	
   reengineering	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   into	
   a	
  
rhodopsin	
  protein	
  mimic,	
  J	
  Am	
  Chem	
  Soc	
  129,	
  6140-­‐6148.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  J	
  Am	
  Chem	
  Soc	
  135,	
  16111-­‐16119.	
  
Wang,	
  W.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2014)	
  The	
  photochemical	
  determinants	
  of	
  color	
  
vision:	
   revealing	
   how	
   opsins	
   tune	
   their	
   chromophore's	
   absorption	
   wavelength,	
  
BioEssays	
  :	
  news	
  and	
  reviews	
  in	
  molecular,	
  cellular	
  and	
  developmental	
  biology	
  36,	
  65-­‐74.	
  
Wang,	
   W.,	
   Nossoni,	
   Z.,	
   Berbasova,	
   T.,	
   Watson,	
   C.	
   T.,	
   Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Vasileiou,	
   C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   electronic	
   absorption	
   of	
   protein-­‐
embedded	
  all-­‐trans-­‐retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
89	
  

48.	
  
	
  
49.	
  
	
  
50.	
  
	
  
51.	
  
	
  
52.	
  

	
  
53.	
  
	
  
54.	
  

55.	
  
	
  
56.	
  

	
  
57.	
  
	
  
58.	
  

Nosrati,	
   M.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   A	
  
Photoisomerizing	
  Rhodopsin	
  Mimic	
  Observed	
  at	
  Atomic	
  Resolution,	
  J	
  Am	
  Chem	
  Soc	
  138,	
  
8802-­‐8808.	
  
Berbasova,	
   T.,	
   Santos,	
   E.	
   M.,	
   Nosrati,	
   M.,	
   Vasileiou,	
   C.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2016)	
  Light-­‐Activated	
  Reversible	
  Imine	
  Isomerization:	
  Towards	
  a	
  Photochromic	
  Protein	
  
Switch,	
  Chembiochem	
  :	
  a	
  European	
  journal	
  of	
  chemical	
  biology	
  17,	
  407-­‐414.	
  
Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Berbasova,	
   T.,	
   Nosrati,	
   M.,	
   Jia,	
   X.,	
   Vasileiou,	
   C.,	
   Wang,	
   W.,	
   Santos,	
   E.	
  
M.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2015)	
  "Turn-­‐on"	
  protein	
  fluorescence:	
  in	
  situ	
  formation	
  
of	
  cyanine	
  dyes,	
  J	
  Am	
  Chem	
  Soc	
  137,	
  1073-­‐1080.	
  
Wang,	
  W.,	
  Nossoni,	
  Z.,	
  Berbasova,	
  T.,	
  Watson,	
  C.	
  T.,	
  Yapici,	
  I.,	
  Lee,	
  K.	
  S.	
  S.,	
  Vasileiou,	
  C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   Electronic	
   Absorption	
   of	
   Protein-­‐
Embedded	
  All-­‐trans-­‐Retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
Folli,	
   C.,	
   Calderone,	
   V.,	
   Ottonello,	
   S.,	
   Bolchi,	
   A.,	
   Zanotti,	
   G.,	
   Stoppini,	
   M.,	
   and	
   Berni,	
   R.	
  
(2001)	
   Identification,	
   retinoid	
   binding,	
   and	
   x-­‐ray	
   analysis	
   of	
   a	
   human	
   retinol-­‐binding	
  
protein,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  98,	
  3710-­‐3715.	
  
Gunasekaran,	
   K.,	
   Hagler,	
   A.	
   T.,	
   and	
   Gierasch,	
   L.	
   M.	
   (2004)	
   Sequence	
   and	
   structural	
  
analysis	
   of	
   cellular	
   retinoic	
   acid-­‐binding	
   proteins	
   reveals	
   a	
   network	
   of	
   conserved	
  
hydrophobic	
  interactions,	
  Proteins-­‐Structure	
  Function	
  and	
  Genetics	
  54,	
  179-­‐194.	
  
Nossoni,	
  Z.,	
  Ass	
  	
  	
  	
  	
  	
  ar,	
  Z.,	
  Yapici,	
  I.,	
  Nosrati,	
  M.,	
  Wang,	
  W.,	
  Berbasova,	
  T.,	
  Vasileiou,	
  C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  Crystallogr.	
  Sect.	
  D.	
  Biol.	
  
Crystallogr.	
  70,	
  3226-­‐3232.	
  
Shameer,	
   K.,	
   Shingate,	
   P.	
   N.,	
   Manjunath,	
   S.	
   C.	
   P.,	
   Karthika,	
   M.,	
   Pugalenthi,	
   G.,	
   and	
  
Sowdhamini,	
   R.	
   (2011)	
   3DSwap:	
   curated	
   knowledgebase	
   of	
   proteins	
   involved	
   in	
   3D	
  
domain	
  swapping,	
  Database-­‐the	
  Journal	
  of	
  Biological	
  Databases	
  and	
  Curation.	
  
Sillitoe,	
   I.,	
   Cuff,	
   A.	
   L.,	
   Dessailly,	
   B.	
   H.,	
   Dawson,	
   N.	
   L.,	
   Furnham,	
   N.,	
   Lee,	
   D.,	
   Lees,	
   J.	
   G.,	
  
Lewis,	
   T.	
   E.,	
   Studer,	
   R.	
   A.,	
   Rentzsch,	
   R.,	
   Yeats,	
   C.,	
   Thornton,	
   J.	
   M.,	
   and	
   Orengo,	
   C.	
   A.	
  
(2013)	
  New	
  functional	
  families	
  (FunFams)	
  in	
  CATH	
  to	
  improve	
  the	
  mapping	
  of	
  conserved	
  
functional	
  sites	
  to	
  3D	
  structures,	
  Nucleic	
  Acids	
  Research	
  41,	
  D490-­‐D498.	
  
Englander,	
   S.	
   W.,	
   and	
   Mayne,	
   L.	
   (2014)	
   The	
   nature	
   of	
   protein	
   folding	
   pathways,	
   Proc	
  
Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  111,	
  15873-­‐15880.	
  
Englander,	
  S.	
  W.,	
  Mayne,	
  L.,	
  and	
  Krishna,	
  M.	
  M.	
  (2007)	
  Protein	
  folding	
  and	
  misfolding:	
  
mechanism	
  and	
  principles,	
  Quarterly	
  reviews	
  of	
  biophysics	
  40,	
  287-­‐326.	
  

	
  
90	
  

59.	
  
	
  
60.	
  
	
  
61.	
  
	
  
62.	
  
	
  
63.	
  
	
  
64.	
  
	
  
65.	
  
	
  
66.	
  
	
  
67.	
  

	
  
68.	
  
	
  
69.	
  
	
  
70.	
  

Krishna,	
   M.	
   M.,	
   and	
   Englander,	
   S.	
   W.	
   (2007)	
   A	
   unified	
   mechanism	
   for	
   protein	
   folding:	
  
predetermined	
   pathways	
   with	
   optional	
   errors,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  16,	
  449-­‐464.	
  
Mayne,	
   L.,	
   and	
   Englander,	
   S.	
   W.	
   (2000)	
   Two-­‐state	
   vs.	
   multistate	
   protein	
   unfolding	
  
studied	
   by	
   optical	
   melting	
   and	
   hydrogen	
   exchange,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  9,	
  1873-­‐1877.	
  
Roder,	
  H.,	
  Elove,	
  G.	
  A.,	
  and	
  Englander,	
  S.	
  W.	
  (1988)	
  Structural	
  characterization	
  of	
  folding	
  
intermediates	
   in	
   cytochrome	
   c	
   by	
   H-­‐exchange	
   labelling	
   and	
   proton	
   NMR,	
   Nature	
   335,	
  
700-­‐704.	
  
Martins,	
   N.	
   F.,	
   and	
   Santoro,	
   M.	
   M.	
   (1999)	
   Partially	
   folded	
   intermediates	
   during	
  
trypsinogen	
  denaturation,	
  Brazilian	
  journal	
  of	
  medical	
  and	
  biological	
  research	
  =	
  Revista	
  
brasileira	
  de	
  pesquisas	
  medicas	
  e	
  biologicas	
  32,	
  673-­‐682.	
  
Medina,	
   E.,	
   Cordova,	
   C.,	
   Villalobos,	
   P.,	
   Reyes,	
   J.,	
   Komives,	
   E.	
   A.,	
   Ramirez-­‐Sarmiento,	
   C.	
  
A.,	
   and	
   Babul,	
   J.	
   (2016)	
   Three-­‐Dimensional	
   Domain	
   Swapping	
   Changes	
   the	
   Folding	
  
Mechanism	
  of	
  the	
  Forkhead	
  Domain	
  of	
  FoxP1,	
  Biophysical	
  journal	
  110,	
  2349-­‐2360.	
  
Zwanzig,	
   R.	
   (1997)	
   Two-­‐state	
   models	
   of	
   protein	
   folding	
   kinetics,	
   Proceedings	
   of	
   the	
  
National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  94,	
  148-­‐150.	
  
Sanson,	
   B.,	
   Wang,	
   T.,	
   Sun,	
   J.,	
   Wang,	
   L.,	
   Kaczocha,	
   M.,	
   Ojima,	
   I.,	
   Deutsch,	
   D.,	
   and	
   Li,	
   H.	
  
(2014)	
  Crystallographic	
  study	
  of	
  FABP5	
  as	
  an	
  intracellular	
  endocannabinoid	
  transporter,	
  
Acta	
  crystallographica.	
  Section	
  D,	
  Biological	
  crystallography	
  70,	
  290-­‐298.	
  
Assar,	
   Z.,	
   Nossoni,	
   Z.,	
   Wang,	
   W.,	
   Santos,	
   Elizabeth	
  M.,	
   Kramer,	
   K.,	
   McCornack,	
   C.,	
  
Vasileiou,	
  C.,	
  Borhan,	
  B.,	
  and	
  Geiger,	
  James	
  H.	
  Domain-­‐Swapped	
  Dimers	
  of	
  Intracellular	
  
Lipid-­‐Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure.	
  
Silvaroli,	
   J.	
   A.,	
   Arne,	
   J.	
   M.,	
   Chelstowska,	
   S.,	
   Kiser,	
   P.	
   D.,	
   Banerjee,	
   S.,	
   and	
   Golczak,	
   M.	
  
(2016)	
   Ligand	
   Binding	
   Induces	
   Conformational	
   Changes	
   in	
   Human	
   Cellular	
   Retinol-­‐
binding	
  Protein	
  1	
  (CRBP1)	
  Revealed	
  by	
  Atomic	
  Resolution	
  Crystal	
  Structures,	
  The	
  Journal	
  
of	
  biological	
  chemistry	
  291,	
  8528-­‐8540.	
  
Gillilan,	
  R.	
  E.,	
  Ayers,	
  S.	
  D.,	
  and	
  Noy,	
  N.	
  (2007)	
  Structural	
  basis	
  for	
  activation	
  of	
  fatty	
  acid-­‐
binding	
  protein	
  4,	
  J	
  Mol	
  Biol	
  372,	
  1246-­‐1260.	
  
Sessler,	
  R.	
  J.,	
  and	
  Noy,	
  N.	
  (2005)	
  A	
  ligand-­‐activated	
  nuclear	
  localization	
  signal	
  in	
  cellular	
  
retinoic	
  acid	
  binding	
  protein-­‐II,	
  Molecular	
  cell	
  18,	
  343-­‐353.	
  
Armstrong,	
  E.	
  H.,	
  Goswami,	
  D.,	
  Griffin,	
  P.	
  R.,	
  Noy,	
  N.,	
  and	
  Ortlund,	
  E.	
  A.	
  (2014)	
  Structural	
  
Basis	
   for	
   Ligand	
   Regulation	
   of	
   the	
   Fatty	
   Acid-­‐binding	
   Protein	
   5,	
   Peroxisome	
   Proliferator-­‐
91	
  

	
  
71.	
  
	
  
72.	
  

	
  
73.	
  
	
  
74.	
  
	
  
75.	
  
	
  
76.	
  

	
  
77.	
  
	
  
78.	
  
	
  
79.	
  
	
  
80.	
  

activated	
   Receptor	
   β/δ	
   (FABP5-­‐PPARβ/δ)	
   Signaling	
   Pathway,	
   The	
   Journal	
   of	
   biological	
  
chemistry	
  289,	
  14941-­‐14954.	
  
Greenfield,	
  N.	
  J.	
  (2006)	
  Using	
  circular	
  dichroism	
  collected	
  as	
  a	
  function	
  of	
  temperature	
  
to	
  determine	
  the	
  thermodynamics	
  of	
  protein	
  unfolding	
  and	
  binding	
  interactions,	
  Nature	
  
Protocols	
  1,	
  2527-­‐2535.	
  
Ramos,	
   C.	
   R.	
   R.,	
   Oyama,	
   S.,	
   Jr.,	
   Sforca,	
   M.	
   L.,	
   Pertinhez,	
   T.	
   A.,	
   Ho,	
   P.	
   L.,	
   and	
   Spisni,	
   A.	
  
(2008)	
   2POA:	
   Schistosoma	
   mansoni	
   Sm14	
   Fatty	
   Acid-­‐Binding	
   Protein:	
   improvement	
   of	
  
protein	
   stability	
   by	
   substitution	
   of	
   the	
   single	
   Cys62	
   residue,	
   Worldwide	
   Protein	
   Data	
  
Bank.	
  
Gill,	
   S.	
   C.,	
   and	
   Vonhippel,	
   P.	
   H.	
   (1989)	
   CALCULATION	
   OF	
   PROTEIN	
   EXTINCTION	
  
COEFFICIENTS	
   FROM	
   AMINO-­‐ACID	
   SEQUENCE	
   DATA,	
   Analytical	
   Biochemistry	
   182,	
   319-­‐
326.	
  
Otwinowski,	
   Z.,	
   and	
   Minor,	
   W.	
   (1997)	
   Processing	
   of	
   X-­‐ray	
   diffraction	
   data	
   collected	
   in	
  
oscillation	
  mode,	
  Macromolecular	
  Crystallography,	
  Pt	
  A	
  276,	
  307-­‐326.	
  
Tarter,	
  M.,	
  Capaldi,	
  S.,	
  Carrizo,	
  M.	
  E.,	
  Ambrosi,	
  E.,	
  Perduca,	
  M.,	
  and	
  Monaco,	
  H.	
  L.	
  (2008)	
  
Crystal	
   structure	
   of	
   human	
   cellular	
   retinol-­‐binding	
   protein	
   11	
   to	
   1.2	
   angstrom	
  
resolution,	
  Proteins-­‐Structure	
  Function	
  and	
  Bioinformatics	
  70,	
  1626-­‐1630.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Davis,	
  I.	
  W.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  
J.,	
   Hung,	
   L.	
   W.,	
   Kapral,	
   G.	
   J.,	
   Grosse-­‐Kunstleve,	
   R.	
   W.,	
   McCoy,	
   A.	
   J.,	
   Moriarty,	
   N.	
   W.,	
  
Oeffner,	
  R.,	
  Read,	
  R.	
  J.,	
  Richardson,	
  D.	
  C.,	
  Richardson,	
  J.	
  S.,	
  Terwilliger,	
  T.	
  C.,	
  and	
  Zwart,	
  
P.	
   H.	
   (2010)	
   PHENIX:	
   a	
   comprehensive	
   Python-­‐based	
   system	
   for	
   macromolecular	
  
structure	
  solution,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  213-­‐
221.	
  
Bailey,	
  S.	
  (1994)	
  THE	
  CCP4	
  SUITE	
  -­‐	
  PROGRAMS	
  FOR	
  PROTEIN	
  CRYSTALLOGRAPHY,	
  Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  50,	
  760-­‐763.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (1997)	
   MOLREP:	
   an	
   automated	
   program	
   for	
   molecular	
  
replacement,	
  Journal	
  of	
  Applied	
  Crystallography	
  30,	
  1022-­‐1025.	
  
Emsley,	
   P.,	
   Lohkamp,	
   B.,	
   Scott,	
   W.	
   G.,	
   and	
   Cowtan,	
   K.	
   (2010)	
   Features	
   and	
   development	
  
of	
  Coot,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  486-­‐501.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  J.,	
  Hung,	
  L.-­‐
W.,	
  Jain,	
  S.,	
  Kapral,	
  G.	
  J.,	
  Kunstleve,	
  R.	
  W.	
  G.,	
  McCoy,	
  A.	
  J.,	
  Moriarty,	
  N.	
  W.,	
  Oeffner,	
  R.	
  
D.,	
   Read,	
   R.	
   J.,	
   Richardson,	
   D.	
   C.,	
   Richardson,	
   J.	
   S.,	
   Terwilliger,	
   T.	
   C.,	
   and	
   Zwart,	
   P.	
   H.	
  
(2011)	
  The	
  Phenix	
  software	
  for	
  automated	
  determination	
  of	
  macromolecular	
  structures,	
  
Methods	
  55,	
  94-­‐106.	
  
92	
  

81.	
  
	
  
71.	
  
	
  
72.	
  

	
  
73.	
  
	
  
74.	
  
	
  
75.	
  
	
  
76.	
  

	
  
77.	
  
	
  
78.	
  
	
  
79.	
  
	
  
80.	
  

Murshudov,	
  G.	
  N.,	
  Vagin,	
  A.	
  A.,	
  and	
  Dodson,	
  E.	
  J.	
  (1997)	
  Refinement	
  of	
  macromolecular	
  
structures	
   by	
   the	
   maximum-­‐likelihood	
   method,	
   Acta	
   Crystallographica	
   Section	
   D-­‐
Biological	
  Crystallography	
  53,	
  240-­‐255.	
  
Greenfield,	
  N.	
  J.	
  (2006)	
  Using	
  circular	
  dichroism	
  collected	
  as	
  a	
  function	
  of	
  temperature	
  
to	
  determine	
  the	
  thermodynamics	
  of	
  protein	
  unfolding	
  and	
  binding	
  interactions,	
  Nature	
  
Protocols	
  1,	
  2527-­‐2535.	
  
Ramos,	
   C.	
   R.	
   R.,	
   Oyama,	
   S.,	
   Jr.,	
   Sforca,	
   M.	
   L.,	
   Pertinhez,	
   T.	
   A.,	
   Ho,	
   P.	
   L.,	
   and	
   Spisni,	
   A.	
  
(2008)	
   2POA:	
   Schistosoma	
   mansoni	
   Sm14	
   Fatty	
   Acid-­‐Binding	
   Protein:	
   improvement	
   of	
  
protein	
   stability	
   by	
   substitution	
   of	
   the	
   single	
   Cys62	
   residue,	
   Worldwide	
   Protein	
   Data	
  
Bank.	
  
Gill,	
   S.	
   C.,	
   and	
   Vonhippel,	
   P.	
   H.	
   (1989)	
   CALCULATION	
   OF	
   PROTEIN	
   EXTINCTION	
  
COEFFICIENTS	
   FROM	
   AMINO-­‐ACID	
   SEQUENCE	
   DATA,	
   Analytical	
   Biochemistry	
   182,	
   319-­‐
326.	
  
Otwinowski,	
   Z.,	
   and	
   Minor,	
   W.	
   (1997)	
   Processing	
   of	
   X-­‐ray	
   diffraction	
   data	
   collected	
   in	
  
oscillation	
  mode,	
  Macromolecular	
  Crystallography,	
  Pt	
  A	
  276,	
  307-­‐326.	
  
Tarter,	
  M.,	
  Capaldi,	
  S.,	
  Carrizo,	
  M.	
  E.,	
  Ambrosi,	
  E.,	
  Perduca,	
  M.,	
  and	
  Monaco,	
  H.	
  L.	
  (2008)	
  
Crystal	
   structure	
   of	
   human	
   cellular	
   retinol-­‐binding	
   protein	
   11	
   to	
   1.2	
   angstrom	
  
resolution,	
  Proteins-­‐Structure	
  Function	
  and	
  Bioinformatics	
  70,	
  1626-­‐1630.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Davis,	
  I.	
  W.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  
J.,	
   Hung,	
   L.	
   W.,	
   Kapral,	
   G.	
   J.,	
   Grosse-­‐Kunstleve,	
   R.	
   W.,	
   McCoy,	
   A.	
   J.,	
   Moriarty,	
   N.	
   W.,	
  
Oeffner,	
  R.,	
  Read,	
  R.	
  J.,	
  Richardson,	
  D.	
  C.,	
  Richardson,	
  J.	
  S.,	
  Terwilliger,	
  T.	
  C.,	
  and	
  Zwart,	
  
P.	
   H.	
   (2010)	
   PHENIX:	
   a	
   comprehensive	
   Python-­‐based	
   system	
   for	
   macromolecular	
  
structure	
  solution,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  213-­‐
221.	
  
Bailey,	
  S.	
  (1994)	
  THE	
  CCP4	
  SUITE	
  -­‐	
  PROGRAMS	
  FOR	
  PROTEIN	
  CRYSTALLOGRAPHY,	
  Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  50,	
  760-­‐763.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (1997)	
   MOLREP:	
   an	
   automated	
   program	
   for	
   molecular	
  
replacement,	
  Journal	
  of	
  Applied	
  Crystallography	
  30,	
  1022-­‐1025.	
  
Emsley,	
   P.,	
   Lohkamp,	
   B.,	
   Scott,	
   W.	
   G.,	
   and	
   Cowtan,	
   K.	
   (2010)	
   Features	
   and	
   development	
  
of	
  Coot,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  486-­‐501.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  J.,	
  Hung,	
  L.-­‐
W.,	
  Jain,	
  S.,	
  Kapral,	
  G.	
  J.,	
  Kunstleve,	
  R.	
  W.	
  G.,	
  McCoy,	
  A.	
  J.,	
  Moriarty,	
  N.	
  W.,	
  Oeffner,	
  R.	
  
D.,	
   Read,	
   R.	
   J.,	
   Richardson,	
   D.	
   C.,	
   Richardson,	
   J.	
   S.,	
   Terwilliger,	
   T.	
   C.,	
   and	
   Zwart,	
   P.	
   H.	
  

93	
  

	
  
81.	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

(2011)	
  The	
  Phenix	
  software	
  for	
  automated	
  determination	
  of	
  macromolecular	
  structures,	
  
Methods	
  55,	
  94-­‐106.	
  
Murshudov,	
  G.	
  N.,	
  Vagin,	
  A.	
  A.,	
  and	
  Dodson,	
  E.	
  J.	
  (1997)	
  Refinement	
  of	
  macromolecular	
  
structures	
   by	
   the	
   maximum-­‐likelihood	
   method,	
   Acta	
   Crystallographica	
   Section	
   D-­‐
Biological	
  Crystallography	
  53,	
  240-­‐255.	
  

	
  
	
  
	
  
	
  
	
  
	
  
94	
  

Chapter	
  III:	
  Can	
  we	
  predict	
  and	
  control	
  the	
  folding	
  product	
  of	
  proteins	
  through	
  their	
  
amino	
  acid	
  sequence	
  	
  (at	
  least	
  in	
  iLBP	
  family)?	
  
	
  

III-­‐1	
  Introduction	
  
	
  

As	
  we	
  discussed	
  in	
  the	
  previous	
  chapter,	
  intracellular	
  lipid	
  binding	
  proteins	
  (iLBP)	
  are	
  a	
  

family	
   of	
   relatively	
   small,	
   soluble	
   proteins.	
   They	
   chaperone	
   lypophilic,	
   insoluble	
   molecules	
  
through	
   the	
   cytosol	
   of	
   virtually	
   all	
   animal	
   tissues.(106,	
  107)	
   Fatty	
   acid	
   binding	
   proteins	
   (FABPs)	
   are	
  
members	
   of	
   this	
   family(107).	
   Recently	
   it	
   was	
   shown	
   that	
   the	
   inhibition	
   of	
   intracellular	
  
endocannabinoid	
   carriers	
   such	
   as	
   FABPs	
   may	
   provide	
   an	
   alternative	
   strategy	
   to	
   modulate	
  
endocannabinoid	
   inactivation	
   and	
   anti-­‐inflammatory	
   effects.	
   (93)	
   There	
   is	
   also	
   evidence	
   of	
   the	
  
direct	
   transfer	
   of	
   their	
   lypophilic	
   cargo	
   to	
   the	
   regulatory	
   transcription	
   factors,	
   including	
   the	
  
retinoic	
   acid	
   receptor	
   (RAR)	
   in	
   the	
   case	
   of	
   retinoic	
   acid	
   and	
   the	
   peroxisome	
   proliferator-­‐
activated	
   receptor	
   (PPAR)	
   in	
   the	
   case	
   of	
   Linoleic	
   acid.	
   (108)	
   Many	
   other	
   recent	
   reports	
   have	
  
suggested	
  a	
  mechanism	
  of	
  allosteric	
  regulation	
  for	
  some	
  iLBP	
  family	
  members,	
  namely	
  FABP4	
  
and	
  CRABPII.	
  (96)	
  (97,	
  98)	
  Specifically,	
  it	
  has	
  been	
  suggested	
  in	
  the	
  case	
  of	
  both	
  FABP4	
  and	
  CRABPII	
  
that	
  ligand	
  binding	
  leads	
  to	
  conformational	
  change	
  that	
  results	
  in	
  translocation	
  to	
  the	
  nucleus,	
  
delivering	
  either	
  retinoic	
  acid	
  or	
  fatty	
  acids	
  to	
  their	
  respective	
  nuclear	
  receptors.	
  (96)	
  (97,	
  98)	
  
However,	
   structures	
   of	
   apo	
   and	
   holo	
   iLBP	
   monomers	
   show	
   only	
   subtle	
   changes	
   in	
  
structure	
   upon	
   binding.	
   Based	
   on	
   our	
   results,	
   discussed	
   in	
   section	
   III-­‐6,	
   we	
   suggest	
   the	
   DS	
  
dimer	
   structure	
   to	
   be	
   exquisitely	
   sensitive	
   to	
   ligand	
   binding,	
   undergoing	
   reproducible	
   and	
  
dramatic	
   conformational	
   change	
   upon	
   ligand	
   association.	
   This	
   begs	
   the	
   question	
   of	
   whether	
  
the	
   DS	
   dimer	
   may	
   form	
   in	
   vivo	
   for	
   some	
   members	
   of	
   the	
   iLBP	
   family,	
   and	
   play	
   important	
  
physiological	
  roles,	
  especially	
  related	
  to	
  allosteric	
  regulation.	
  
95	
  

The	
  ultimate	
  goal	
  of	
  our	
  research	
  is	
  to	
  have	
  the	
  tools	
  necessary	
  to	
  predict	
  the	
  folding	
  
products	
  of	
  iLBP	
  family	
  members	
  based	
  on	
  their	
  sequences.	
  This	
  will	
  require	
  understanding	
  of	
  
the	
  effect	
  of	
  sequence	
  insertions,	
  significant	
  changes	
  in	
  amino	
  acid	
  identity	
  at	
  certain	
  positions,	
  
etc.	
  This	
  would	
  	
  allow	
  us	
  to	
  control	
  and	
  predict	
  the	
  folding	
  pathway	
  of	
  iLBP	
  family	
  members	
  to	
  
achieve	
  the	
  desired	
  folding	
  product,	
  either	
  monomer	
  or	
  DS	
  dimer.	
  The	
  accomplishment	
  of	
  this	
  
goal	
   will	
   allow	
   us	
   to	
   predict	
   from	
   sequence	
   the	
   likely	
   folding	
   outcome	
   for	
   members	
   of	
   the	
  
family,	
  and	
  will	
  also	
  allow	
  us	
  to	
  control	
  at	
  will	
  the	
  folding	
  product	
  via	
  mutagenesis	
  

III-­‐2	
  Investigation	
  of	
  potential	
  domain	
  swapping	
  in	
  other	
  iLBP	
  family	
  members:	
  Case	
  
Study	
  1,	
  human	
  FABP4	
  
III-­‐2-­‐1	
  History	
  of	
  FABP4	
  
As	
   previously	
   mentioned,	
   gene-­‐activating	
   lipophilic	
   compounds	
   are	
   carried	
   from	
   the	
  
cytosol	
   to	
   the	
   nucleus	
   by	
   FABP.	
   	
   FABP4	
   (adipocyte	
   FABP),	
   and	
   FABP5	
   (epidermal	
   FABP)	
  
specifically	
   deliver	
   to	
   Peroxisome-­‐proliferators-­‐activated	
   receptors	
   (PPARγ,	
   and	
   PPARβ/δ	
  
repectively)	
  followed	
  by	
  enhancing	
  the	
  transcriptional	
  activity	
  of	
  the	
  receptor.	
  	
  
	
  
	
  

Lid	
  

	
  
	
  
	
  
	
  

	
  
Figure	
   III-­‐1:	
   Overlaid	
   of	
   FABP4	
   (2q9s,	
   green)	
   -­‐Linoleic	
   acid	
   (blue),	
   FABP4	
   (2ans,	
   pink)-­‐oleic	
  
acid	
  (grey)	
  shows	
  the	
  exit	
  of	
  the	
  oleic	
  acid	
  from	
  the	
  binding	
  pocket.	
  

96	
  

The	
   activated	
   receptor	
   is	
   involved	
   in	
   regulation	
   of	
   adipocyte	
   differentiation,	
   insulin	
  
sensitivity	
   etc.	
   (96)	
   (97,	
  98)	
   FABP4	
   binds	
   to	
   	
   activating	
   (linoleic	
   acid)	
   and	
   non-­‐activating	
   compounds	
  
(oleic	
  acid).	
  	
  Recent	
  studies	
  showed	
  activated	
  and	
  non-­‐activated	
  conformations	
  of	
  the	
  protein	
  
have	
  different	
  homodimeric	
  configurations	
  through	
  crystal	
  packing.	
  Non-­‐activating	
  ligands	
  exit	
  
from	
  the	
  portal	
  (lid),	
  preventing	
  binding	
  pocket	
  closure.	
  (Figure	
   III-­‐1)	
  Crystallographic	
  symmetry	
  
yields	
  a	
  homodimer	
  in	
  which	
  opposing	
  helix-­‐turn-­‐helix	
  subdomains	
  form	
  the	
  principal	
  protein–
protein	
  contact	
  surface	
  followed	
  by	
  masking	
  its	
  NLS	
  site.	
  44	
  	
  

III-­‐2-­‐2	
  Investigation	
  on	
  hFABP4	
  bacterial	
  expression	
  
Initially	
   we	
   picked	
   this	
   protein	
   as	
   the	
   first	
   case	
   study.	
   We	
   thought	
   the	
   “homodimeric	
  
configuration”	
   due	
   to	
   crystal	
   packing	
   might	
   be	
   the	
   incorrectly	
   refined	
   DS	
   dimer	
   structure.	
  
subsequently	
   we	
   found	
   out	
   about	
   the	
   presence	
   of	
   dimer	
   species	
   in	
   our	
   size	
   exclusion	
  
chromatography	
  profile.	
  (Figure	
  III-­‐2)	
  However	
  it	
  was	
  due	
  to	
  a	
  disulfide	
  bond	
  linkage	
  between	
  
Cys	
  in	
  the	
  N-­‐terminus	
  of	
  FABP4	
  and	
  it	
  was	
  confirmed	
  by	
  addition	
  of	
  denaturant,	
  2mM	
  DTT,	
  in	
  
our	
   purification	
   buffers	
   and	
   running	
   an	
   SDS-­‐gel	
   with	
   and	
   without	
   fresh	
   denaturant	
   reagent,	
  
which	
   we	
   only	
   saw	
   one	
   species	
   as	
   monomer	
   size	
   whenever	
   DTT	
   was	
   present.	
   Even	
   while	
  
performing	
   denaturation/	
   renaturation	
   experiment	
   (explained	
   in	
   section	
   III-­‐3),	
   only	
   monomer	
  
was	
  obtained	
  at	
  higher	
  concentrations	
  of	
  material	
  (0.3mM).	
  (Data	
  not	
  shown)	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

97	
  

	
  
	
  
	
  
	
  

Monomer

Dimer

	
  
	
  

Figure	
   III-­‐2:	
  Size	
  Exclusion	
  Chromatogram	
  of	
  human	
  FABP4,	
  which	
  was	
  expressed	
  in	
  E.coli	
  (at	
  
30°C	
   as	
   dimer	
   favoring	
   temperature,	
   explained	
   in	
   section	
   III-­‐3,	
   green).	
   SEC	
   profile	
   of	
  
monomer	
  hCRBPII	
  mutant	
  (orange)	
  as	
  our	
  size	
  control.	
  

III-­‐3	
  Investigation	
  of	
  potential	
  domain	
  swapping	
  in	
  other	
  iLBP	
  family	
  members:	
  Case	
  Study	
  1,	
  
human	
  FABP5	
  
III-­‐3-­‐1	
  History	
  of	
  hFABP5	
  
Fatty	
   acid	
   binding	
   protein	
   5	
   (FABP5)	
   is	
   found	
   mostly	
   in	
   epidermal	
   cells,	
   but	
   also	
   in	
   other	
  
	
  
tissues,	
  such	
  as	
  brain,	
  liver,	
  kidney,	
  lung,	
  and	
  adipose	
  tissue. 	
  
Recently	
  it	
  has	
  been	
  shown	
  that	
  FABP5	
  is	
  an	
  intracellular	
  carrier	
  of	
  anandamide	
  (AEA,	
  the	
  best	
  
characterized	
   endocannabinoid	
   which	
   activates	
   cannabinoid	
   receptors	
   in	
   the	
   central	
   nervous	
  
system	
   and	
   peripheral	
   tissues,	
   i.e.	
   CB1)	
   from	
   plasma	
   membrane	
   to	
   	
   intracellular	
   fatty	
   acid	
  
amide	
   hydrolase	
   (FAAH)	
   for	
   hydrolysis,	
   since	
   the	
   period	
   of	
   endocannabinoid	
   signaling,	
   by	
  
endocannabinoid,	
   is	
   regulated	
   through	
   ‘on	
   demand’	
   biosynthesis	
   and	
   prompt	
   catabolism.	
   (93)	
  	
  
(Figure	
  III-­‐3)	
  

98	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

Figure	
  III-­‐3:	
  Model	
  for	
  the	
  role	
  of-­‐FABP5	
  and	
  its	
  involvement	
  in	
  regulating.	
  

	
  
Also	
   it	
   has	
   been	
   shown	
   that	
   the	
   selective	
   FABP5	
   inhibitor,	
   BMS-­‐309403,	
   reduces	
   AEA	
  

inactivation	
   in	
   cells	
   and	
   increases	
   endocannabinoid	
   levels,	
   produceing	
   beneficial	
   anti-­‐
	
  
inflammatory	
  and	
  antinociceptive	
  effects.(109,	
  110) FABP5	
  has	
  been	
  discovered	
  to	
  form	
  a	
  domain	
  
swapped	
  dimer,	
  with	
  hinge	
  residues	
  of	
  Ser59,	
  Thr60,	
  Val61	
  (equivalent	
  of	
  Ser55,	
  Thr56,	
  Phe57	
  
in	
   hCRBPII)	
   (Figure	
   III-­‐4	
   and	
   III-­‐5b).(93)	
  	
   Existence	
   of	
   domain	
   swapping	
   in	
   FABP5	
   was	
   very	
  
interesting	
  for	
  us,	
  since	
  this	
  protein	
  is	
  in	
  the	
  same	
  family	
  as	
  hCRBPII.	
  
	
  
	
  
	
  
	
  
	
  

99	
  

	
  
	
  
	
  

	
  

	
  
	
  

	
  
	
  
	
  
	
  

Figure	
   III-­‐4:	
   Overlaid	
   of	
   DS	
   dimers	
   of	
   holo	
   hFABP5	
   and	
   hCRBPII.	
   HCRBPII	
   (purple,	
   PDB	
   ID	
  
4ZH9),	
   BMS-­‐hFABP5	
   (green),	
   AEA-­‐hFABP5	
   (orange).	
   There	
   is	
   a	
   big	
   different	
   orientation	
   of	
  
two	
  domains	
  upon	
  ligand	
  binding.	
  	
  	
  
Two	
   hFABP5	
   DS	
   dimer	
   structures	
   have	
   thus	
   far	
   been	
   determined.	
   Both	
   are	
   ligand-­‐
bound,	
   but	
   the	
   two	
   ligands	
   are	
   distinct,	
   one	
   being	
   the	
   endocannabinoid	
   anandamide,	
   an	
  
actIIIator	
   and	
   the	
   other	
   BMS-­‐309403,	
   a	
   high-­‐affinity	
   inhibitor.	
   (93)	
   A	
   large	
   change	
   in	
   the	
   relatIIIe	
  
orientation	
   of	
   the	
   two	
   domains	
   is	
   seen	
   in	
   the	
   two	
   ligand-­‐bound	
   DS	
   dimer	
   structures,	
  
presumably	
   due	
   to	
   the	
   very	
   different	
   binding	
   modes	
   of	
   the	
   two	
   ligands.	
   (Figure	
   III-­‐5)	
   It	
   is	
  
therefore	
   possible	
   that	
   different	
   ligands	
   can	
   result	
   in	
   distinct	
   protein	
   conformations,	
   which	
  
could	
  then	
  provide	
  distinct,	
  allosterically	
  regulated	
  function,	
  to	
  the	
  protein.	
  Although	
  validation	
  
of	
  such	
  a	
  mechanism	
  requires	
  further	
  exploration	
  of	
  the	
  iLBP	
  family	
  members	
  in	
  vIIIo,	
  the	
  fact	
  
that	
  the	
  wild-­‐type	
  sequence	
  of	
  a	
  second	
  member	
  of	
  the	
  iLBP	
  family	
  also	
  domain	
  swaps	
  in	
  a	
  very	
  

100	
  

similar	
  way	
  to	
  that	
  of	
  hCRBPII,	
  suggests	
  that	
  a	
  significant	
  subset	
  of	
  iLBP	
  family	
  members	
  may	
  
also	
  form	
  DS	
  dimers.	
  
Therefore,	
  We	
  have	
  conducted	
  mutagenesis	
  and	
  structural	
  elucidation	
  on	
  human	
  Fatty	
  
Acid	
  Binding	
  Protein	
  5	
  (hFABP5)	
  to	
  analyze	
  its	
  domain	
  swapping	
  pathway	
  and	
  with	
  the	
  future	
  
goal	
   of	
   predicting	
   potential	
   sequences	
   of	
   FABP’s	
   (or	
   other	
   iLBPs	
   members)	
   to	
   produce	
   DS	
  
dimers	
  as	
  well.	
  For	
  instance,	
  if	
  we	
  think	
  that	
  the	
  presence	
  of	
  hydrophobic	
  residues	
  in	
  positions	
  
that	
   are	
   only	
   inside	
   the	
   binding	
   pocket	
   in	
   the	
   DS	
   dimer	
   lead	
   to	
   significantly	
   increased	
   DS	
  
dimerization,	
   then	
   it	
   needs	
   to	
   be	
   investigated	
   among	
   other	
   iLBP	
   family	
   members	
   whose	
  
sequences	
  have	
  such	
  residues.	
  Several	
  such	
  sequences	
  can	
  already	
  be	
  identified	
  (Figure	
  III-­‐11)	
  
which	
  other	
  members	
  of	
  our	
  group	
  will	
  investigate	
  on	
  them	
  from	
  now	
  on.	
  Eventually	
  this	
  will	
  
lead	
  to	
  a	
  re-­‐evaluation	
  of	
  the	
  structural	
  biology	
  of	
  the	
  iLBP	
  family.	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  	
  
	
  
	
  
	
  
	
  

101	
  

	
  

a.	
  

	
  

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
b.	
  

	
  
	
  

14.4A%
17A%

	
  
	
  
	
  
	
  
	
  

Figure	
   III-­‐5:	
   a.	
   Overlaid	
   of	
   holo	
   DS	
   dimers	
   of	
   BMS-­‐hFABP5	
   (green),	
   AEA-­‐hFABP5	
   (orange).	
  
Dramatic	
  change	
  is	
  seen	
  in	
  the	
  relative	
  orientation	
  of	
  two	
  domains	
  in	
  these	
  DS	
  dimers,	
  which	
  
is	
  induced	
  by	
  different	
  ligands.	
  b.	
  Overlaid	
  of	
  DS	
  dimers	
  of	
  BMS-­‐hFABP5	
  (green),	
  AEA-­‐hFABP5	
  
(orange)	
   and	
   holo	
   Q108K:K40D	
   DS	
   dimer	
   of	
   hCRBPII.	
   AEA,	
   BMS	
   and	
   retinal	
   are	
   shown	
   as	
  
sticks.	
  Between	
  structures	
  of	
  dimer	
   hCRBPII	
  and	
  AEA-­‐hFABP5,	
  there	
  is	
  14Å	
  deviation	
  while	
  
for	
  BMS-­‐hFABP5	
  dimer	
  there	
  is	
  a	
  larger	
  deviation	
  of	
  17Å.	
  
	
  

102	
  

III-­‐3-­‐2	
  Investigate	
  domain	
  swapping	
  in	
  human	
  fatty	
  acid	
  binding	
  protein	
  5	
  	
  
In	
  order	
  to	
  investigate	
  domain	
  swapping	
  in	
  hFABP5,	
  we	
  have	
  decided	
  to	
  first	
  study	
  the	
  
domain	
  swapping	
  of	
  apo	
  protein	
  (no	
  addition	
  of	
  fatty	
  acids).	
  Surprisingly,	
  during	
  its	
  purification,	
  
the	
  chromatogram	
  of	
  the	
  SEC	
  of	
  the	
  elution	
  fractions	
  of	
  post	
  Ni-­‐NTA	
  purification	
  (since	
  it	
  is	
  a	
  
his-­‐tagged	
  protein)	
  showed	
  different	
  results	
  each	
  time	
  for	
  each	
  batch	
  of	
  expression.	
  Most	
  of	
  
the	
  time,	
  we	
  achieved	
  a	
  large	
  monomer	
  peak	
  and	
  very	
  small	
  peaks	
  for	
  the	
  dimer	
  and	
  higher	
  
oligomers.	
  These	
  observations	
  were	
  the	
  same	
  even	
  at	
  different	
  expression	
  temperatures	
  (20°C	
  
and	
  30°C	
  and	
  37°C),	
  which	
  was	
  different	
  than	
  hCRBPII.	
  However,	
  one	
  time	
  we	
  were	
  able	
  to	
  get	
  
a	
  significant	
  amount	
  of	
  dimer	
  (35%)	
  after	
  purification,	
  which	
  was	
  confirmed	
  by	
  SDS-­‐PAGE.	
  
(FigureIII-­‐6a	
  and	
  6b,	
  The	
  reason	
  for	
  getting	
  different	
  results	
  for	
  expression	
  is	
  still	
  not	
  clear	
  for	
  
us.	
  Further,	
  we	
  have	
  found	
  no	
  evidence	
  of	
  monomer/dimer	
  exchange	
  over	
  10	
  days	
  at	
  room	
  
temperature,	
  though	
  it	
  was	
  reported	
  that	
  monomer/dimer	
  exchange	
  might	
  be	
  occurring	
  much	
  
more	
  rapidly.	
  (93)	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

103	
  

	
  

a.	
  

	
  
	
  
	
  

	
  
	
  
b.	
  

Fr.18

c.	
  

Fr.19

Fr.20

Fr.	21
Monomer

	
  

Dimer

	
  

Fr.	21

Fr.18

Fr.19 Fr.20

Other	Impuri/es

15kDa

	
  

	
  

Figure	
   III-­‐6:	
   a.	
   SEC	
   of	
   WT	
   and	
   Q4A	
   mutant	
   FABP5	
   purification	
   by	
   SEC	
   after	
   E.	
   coli	
  
expression.	
  	
  b.	
  SDS-­‐PAGE	
  of	
  SEC	
  fractions	
  18-­‐21	
  of	
  WT-­‐hFABP5.	
  c.	
  	
  Native	
  gel	
  of	
  the	
  same	
  
SEC	
   fractions.	
   It	
   shows	
   that	
   in	
   the	
   dimer	
   peak,	
   both	
   monomer	
   and	
   dimer	
   species	
   are	
  
	
  
present.	
  
However,	
  these	
  studies	
  were	
  conducted	
  after	
  de-­‐lipidation	
  treatment	
  (we	
  did	
  not	
  treat	
  
it	
   with	
   lipidex),	
   which	
   could	
   potentially	
   decrease	
   the	
   barrier	
   between	
   monomer	
   and	
   dimer.	
  
Basically	
  delipidation	
  process	
  is	
  performed	
  by	
  lipidex	
  resin	
  and	
  it	
  is	
  recommended	
  to	
  remove	
  all	
  
the	
   non-­‐specific	
   lipids	
   (which	
   in	
   this	
   case,	
   they	
   may	
   come	
   from	
   E.coli	
   cells).	
   Followed	
   by	
  
crystallization	
   trials	
   of	
   the	
   fractions	
   we	
   thought	
   to	
   be	
   dimer.	
   We	
   were	
   able	
   to	
   crystallize	
   and	
  
determine	
  the	
  structure	
  in	
  P212121	
  space	
  group	
  (Table	
  III-­‐1).	
  However	
  it	
  was	
  not	
  the	
  DS	
  dimer	
  
104	
  

	
  

stricter	
   as	
   we	
   expected	
   it,	
   instead	
   it	
   was	
   two	
   monomers	
   per	
   asymmetric	
   unit.	
   Where	
   in	
   the	
  
crystal	
  packing,	
  two	
  monomers	
  meeting	
  right	
  at	
  the	
  loop	
  of	
  connecting	
  3rd	
  and	
  4th	
  beta	
  strand	
  
(hinge	
  loop	
  in	
  DS	
  dimer	
  structures)	
  where	
  the	
  swapping	
  should	
  occur,	
  but	
  the	
  density	
  for	
  this	
  
part	
   is	
   not	
   very	
   good	
   to	
   build	
   the	
   DS	
   dimer	
   structure.	
   (Figure	
   III-­‐7,	
   Table	
   III-­‐1)	
   This	
   led	
   us	
   to	
  
believe	
  that	
  the	
  corresponded	
  dimer	
  fractions	
  may	
  be	
  a	
  mix	
  of	
  dimer	
  and	
  monomer.	
  This	
  was	
  
then	
   confirmed	
   by	
   native-­‐gel	
   electrophoresis	
   (Figure	
   III-­‐6c).	
   Though	
   the	
   Dimer	
   species	
   is	
  
present	
  in	
  all	
  the	
  fractions,	
  it	
  is	
  only	
  relatively	
  pure	
  in	
  Fraction	
  18.	
  Pooling	
  all	
  the	
  fractions	
  from	
  
18	
  to	
  20	
  gave	
  a	
  mixture	
  of	
  monomer	
  and	
  dimer.	
  This	
  is	
  due	
  to	
  the	
  fact	
  that	
  monomer	
  and	
  dimer	
  
do	
  not	
  separate	
  well	
  on	
  our	
  gel	
  filtration	
  column.	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  
Figure	
   III-­‐7:	
  Overlaid	
  of	
  our	
  current	
  monomer	
  structure	
  (chain	
  A	
  in	
  green	
  and	
  chain	
  B	
  in	
  pink)	
  
with	
  hFABP5	
  monomer	
  (grey,	
  PDB	
   ID	
   4lkp).	
  There	
  is	
  a	
   different	
   conformation	
  between	
  their	
  
3rd	
  and	
  4th	
  beta	
  sheet	
  (hinge	
  loop	
  region	
  in	
  DS	
  dimer),	
  which	
  is	
  pointed	
  by	
  arrow.	
  
Nevertheless,	
  The	
  structure	
  of	
  our	
  monomer	
  has	
  different	
  conformation	
  compare	
  to	
  the	
  
other	
  monomer	
  structures	
  reported	
  between	
  their	
  3rd	
  and	
  4th	
  beta	
  sheet	
  (hinge	
  loop	
  region	
  in	
  
DS	
  dimer	
  structures)(98)	
  (Figure	
  III-­‐7)	
  All	
  of	
  these	
  raise	
  the	
  hypothesis	
  that	
  due	
  to	
  the	
  existence	
  

105	
  

of	
   both	
   monomer	
   and	
   dimer	
   species	
   in	
   the	
   solution,	
   these	
   monomer	
   crystals	
   maybe	
   were	
  
nucleated	
  by	
  dimer	
  species	
  ,	
  therefore	
  we	
  cannot	
  build	
  these	
  structure	
  models	
  as	
  DS	
  dimers.	
  	
  	
  	
  
After	
  that,	
  we	
  expressed	
  hFABP5	
  multiple	
  times	
  and	
  one	
  time	
  we	
  could	
  get	
  dimer	
  peak	
  
again.	
  This	
  time	
  	
  knowing	
  that	
  only	
  fr.	
  18	
  of	
  the	
  SEC	
  run	
  contains	
  dimer	
  species,	
  we	
  separately	
  
pooled	
  and	
  were	
  able	
  to	
  crystallize	
  it	
  (performed	
  by	
  Nona	
  Ehyai).	
  As	
  we	
  expected,	
  	
  we	
  observed	
  
electron	
   density	
   for	
  the	
   continuous	
   hinge	
   loop	
   region	
   and	
   the	
   structure	
   was	
   determined	
   in	
  
P6322	
   space	
   group	
   as	
   symmetric	
   domain	
  swapped	
   dimer.	
   (Figure	
   III-­‐8a)	
   Although	
   the	
   crystal	
  
packing	
   of	
   this	
   structure	
   and	
   the	
   previously	
   mentioned	
   monomer	
   structure	
   are	
   comparable,	
  
there	
  is	
  a	
  large	
  deviation	
  in	
  their	
  relative	
  orientation	
  of	
  the	
  two	
  molecules	
  in	
  those	
  structures	
  
(Figure	
   III-­‐8b)	
   similar	
   to	
   the	
   	
   Figure	
   III-­‐5.	
   Therefor,	
   this	
   may	
   give	
   some	
   evidence	
   that	
   the	
  
previous	
   pseudo	
   monomer	
   structure	
   was	
   nucleated	
   by	
   a	
   dimer.	
   However,	
   there	
   is	
   an	
  extra	
  
electron	
  density	
  in	
  one	
  of	
  the	
  dimer	
  chains	
  binding	
  pocket,	
  which	
  seems	
  to	
  be	
  the	
  size	
  of	
  a	
  fatty	
  
acid.	
   Since	
   we	
   never	
   added	
   any	
   potential	
   ligand	
   to	
   the	
   protein	
   used	
   in	
   crystallization,	
   we	
  
suspect	
  that	
  the	
  ligand	
  is	
  a	
  fatty	
  acid	
  from	
  the	
  E.	
  coli	
  the	
  protein	
  was	
  expressed	
  in.	
  Interestingly,	
  
this	
   DS	
   dimer	
   structure	
   share	
   very	
   similar	
   relative	
   orientation	
   compare	
   to	
   the	
   DS	
   dimer	
  
structure	
  of	
  hFABP5-­‐AEA.	
  (Figure	
  III-­‐8c).	
  
As	
   previously	
   mentioned,	
   the	
   structure	
   of	
   apo	
   hFABP5	
   domain	
   swapped	
  dimer	
   is	
   not	
  
known	
  and	
  is	
  presumably	
  the	
  product	
  of	
  dimerization	
  during	
  folding.	
  Our	
  group	
  actively	
  will	
  try	
  
to	
   obtain	
   the	
   apo	
   structure	
   of	
   DS	
   dimer	
   hFABP5	
   and	
   to	
   use	
   it	
   to	
   design	
   a	
   set	
   of	
   mutagenesis	
  
experiments	
  to	
  probe	
  the	
  determinants	
  of	
  the	
  monomer/dimer	
  ratio	
  in	
  FABP5	
  as	
  described	
  in	
  
chapter	
   III	
   for	
   hCRBPII.	
   A	
   similar	
   strategy	
   will	
   be	
   used	
   to	
   probe	
   the	
   phase	
   relationship	
   in	
   FABP5	
  
as	
   described	
   for	
   hCRBPII	
   as	
   well.	
   (Figure	
   III-­‐11)	
   In	
   order	
   to	
   achieve	
   this	
   structure	
   and	
  

106	
  

understand	
  dimer	
  formation	
  for	
  this	
  protein,	
  the	
  reasons	
  behind	
  the	
  different	
  expression	
  levels	
  
of	
   domain	
  swapped	
   dimers	
   need	
   to	
   be	
   investigated,	
   since	
   the	
   expression	
   process	
  is	
   not	
   the	
  
same	
  every	
  time.	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

a.	
  

b.	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  

c.	
  

Figure	
  III_8:	
  a.	
  Crystal	
  structure	
   of	
  DS	
   dimer	
  of	
  hFABP5	
  of	
  our	
  trial.	
   Chain	
  A	
   in	
  blue,	
  Chain	
  B	
  
in	
  cyan.	
  b.	
  Overlaid	
   of	
   DS	
   hFABP5	
   structure	
   	
   (obtained	
   by	
  Nona,	
  cyan)	
  with	
   abovementioned	
  
pseudo	
  monomer	
  hFABP5	
  (chain	
  A	
  in	
  pink	
  and	
  chain	
  B	
  in	
  green)c.	
   Overlay	
  of	
  our	
  determined	
  
DS	
   hFABP5	
  structure	
  	
  (blue)	
  with	
   DS	
   dimer	
  of	
  BMS-­‐hFABP5	
  (green,	
   4AZM),	
   and	
   AEA-­‐hFABP5	
  
(orange,	
   4AZR).	
   	
   This	
   dimer	
   structure	
   is	
   very	
   similar	
   to	
   the	
   AEA-­‐	
   structure	
   and	
   both	
   have	
  
different	
  orientation	
  of	
  their	
  domains	
  compared	
  with	
  the	
  BMS-­‐structure.	
  	
  
107	
  

III-­‐3-­‐3	
  Study	
  domain	
  swapping	
  occurrence	
  by	
  mutational	
  analysis	
  of	
  human	
  FABP5	
  
We	
  were	
  looking	
  for	
  mutations	
  that	
  could	
  increase	
  the	
  ratio	
  of	
  dimerization	
  of	
  hFABP5.	
  
In	
  order	
  to	
  do	
  this,	
  we	
  compared	
  the	
  structure	
  of	
  the	
  monomer	
  and	
  dimer	
  of	
  hFABP5.	
  In	
  the	
  
monomer,	
   Thr64	
   (equivalent	
   of	
   Y60	
   in	
   hCRBPII,	
   Figure	
   III-­‐11)	
   points	
   in,	
   toward	
   the	
   binding	
  
pocket	
  and	
  makes	
  a	
  water-­‐mediated	
  hydrogen	
  bond	
  with	
  Thr57,	
  (Note:	
  res	
  59	
  to	
  61	
  are	
  in	
  the	
  
hinge	
   loop	
   region).	
   (Figure	
   III-­‐9b)	
   However,	
   in	
   the	
   dimer,	
   Thr64	
   points	
   outside	
   and	
   there	
   is	
   a	
  
new	
  hydrogen	
  bond	
  between	
  of	
  one	
  chain	
  to	
  Thr57	
  of	
  the	
  other	
  chain	
  of	
  the	
  dimer	
  (Figure	
  III-­‐
9a).	
   We	
   reasoned	
   that	
   since	
   both	
   Gln65	
   and	
   Thr64	
   switche	
   sides	
   from	
   outside	
   to	
   inside	
   the	
  
binding	
   pocket	
   in	
   the	
   monomer	
   and	
   DS	
   dimer,	
   respectively,	
   this	
   position	
   likely	
   to	
   be	
   a	
  
“mutational	
  hot	
  spot”	
  for	
  DS	
  dimerization	
  
a.	
  
	
  
	
  

T57	
  
Q65	
  

	
  

3.3	A%

T63	
  

3.0	A%
T64	
  

	
  

b.	
  

	
  
	
  

T64	
  

	
  

3.6	A&

	
  

3.0	A&

T57	
  

	
  
Q65	
  

T63	
  

	
  

Figure	
   III-­‐9:	
   a.	
   In	
  DS	
  dimer	
  of	
  hFABP5	
  (Chain	
  A	
  in	
  salmon,	
  chain	
  B	
  in	
  cyan,	
  PDB	
  4AZR),	
  there	
  is	
  
a	
   hydrogen	
   bond	
   between	
   Q65	
   and	
   T57.	
   b.	
   While	
   in	
   monomer	
   structure	
   (purple,	
   PDB	
   ID	
  
4LKP)	
  Gln65	
   is	
  pointing	
  out	
  of	
  toward	
   the	
  solvent	
  and	
  T64	
  makes	
   a	
  hydrogen	
  bond	
  to	
   Thr57	
  
via	
  water	
  network.	
  Res.	
  57,	
  64	
  and	
  65	
  are	
  represented	
  in	
  sticks.	
  Water	
  molecules	
  are	
  in	
  red.	
  	
  

108	
  

Therefore,	
   we	
   thought	
   that	
   by	
   removing	
   this	
   hydrogen	
   bond	
   between	
   Gln65	
   with	
   Thr57	
  
in	
  the	
  dimer,	
  the	
  formation	
  of	
  the	
  monomer	
  could	
  be	
  more	
  feasible.	
  We	
  mutated	
  Gln	
  to	
  Ala,	
  
which	
   is	
   a	
   small	
   hydrophobic	
   residue.	
   This	
   mutation	
   yielded	
   almost	
   no	
   dimer	
   formation	
  
compare	
   to	
   the	
   WT-­‐hFABP5	
   which	
   confirmed	
   by	
   native	
   gel	
   electrophoreses	
   (Figure	
   IV-­‐6a)	
  
Although,	
  for	
  hFABP5	
  we	
  cannot	
  say	
  exact	
  ratio	
  of	
  monomer/dimer	
  like	
  hCRBPII	
  since	
  we	
  could	
  
not	
   get	
   dimer	
   dependently	
   than	
   monomer.	
   We	
   could	
   get	
   small	
   crystals	
   of	
   this	
   mutant	
   but	
   they	
  
did	
   not	
   diffract.	
   	
   Regarding	
   this	
   scenario,	
   Q65	
   was	
   mutated	
   to	
   W	
   and	
   M	
   as	
   well,	
  (Figure	
  III-­‐10a)	
  
which	
   in	
   both	
   mutations,	
   monomer	
   of	
   hFABP5	
   was	
   only	
   expressed	
   (though	
   Q65M	
   made	
   higher	
  
order	
   oligomers	
   in	
   non	
   reducing	
   environment).	
   We	
   could	
   get	
   only	
   2	
   crystals	
   of	
   the	
   Q65W	
  
mutant,	
  unfortunately	
  they	
  diffracted	
  poorly,	
  to	
  about	
  5Å.	
  
We	
   generated	
   more	
   mutations	
   of	
   hFABP5.	
   However,	
   no	
   mutation	
   has	
   been	
   discovered	
   yet	
   that	
  
increased	
  the	
  ratio	
  of	
  dimer	
  to	
  monomer	
  significantly.	
  	
  
	
  
a.	
  

b.	
  
	
  

	
  
	
  
	
  
	
  
	
  

Figure	
   III-­‐10:	
  a.	
   Size	
  exclusion	
  chromatogram	
  of	
  Gln65	
  mutants	
  of	
  hFABP5	
  during	
  purification	
  
by	
   SEC.	
   b.	
   Size	
   exclusion	
   chromatogram	
   of	
   Thr64	
   mutants	
   of	
   hFABP5	
   (Thr63E	
   pink,	
   T63L	
  
green)	
  during	
  purification	
  by	
  SEC.	
  

109	
  

We	
  applied	
  the	
  phase	
  relationship	
  for	
  monomer	
  and	
  dimer	
  hFABP5	
  (rephrasing	
  occurs	
  
at	
  Phe	
  66.	
  Refer	
  to	
  section	
  II-­‐4-­‐3	
  for	
  more	
  details,	
  Figure	
   III-­‐10b	
   and	
   III-­‐11).	
  For	
  example,	
  the	
  
Thr	
  64	
  side	
  chain	
  is	
  going	
  from	
  inside	
  to	
  outside	
  of	
  the	
  binding	
  cavity	
  in	
  monomer	
  to	
  dimer,	
  it	
  
could	
   be	
   mutated	
   to	
   hydrophilic	
   residues	
   to	
   increase	
   the	
   dimer	
   formation.	
   To	
   test	
   this	
  
hypothesis,	
   we	
   first	
   mutated	
   Thr64	
   to	
   Leu	
   (small	
   hydrophobic	
   residue	
   to	
   mimic	
   the	
   Y60L	
  
mutant	
  of	
  hCRBPII).	
  It	
  did	
  not	
  express	
  well	
  and	
  the	
  yield	
  was	
  low.	
  Then	
  T64E	
  was	
  mutated	
  and	
  
expressed	
  which	
  it	
  was	
  mostly	
  monomer	
  and	
  higher	
  oligomers.	
  (Figure	
  III-­‐10b)	
  Other	
  members	
  
of	
  our	
  group	
  are	
  carrying	
  on	
  the	
  crystallization	
  trial	
  for	
  these	
  mutants.	
  
As	
   our	
   other	
   mutation	
   candidates,	
   Thr63	
   can	
   be	
   chosen	
   to	
   some	
   residues,	
   which	
   are	
  
more	
  hydrophobic.	
  As	
  it	
  mentioned	
  before,	
  hinge	
  loop	
  residues	
  involved	
  in	
  hFABP5	
  DS	
  dimer	
  
structures	
  are:	
  Ser59,	
  Thr60,	
  Val61	
  (the	
  same	
  as	
  Ser55,	
  Thr56	
  and	
  phe57	
  in	
  hCRBPII).	
  Therefore,	
  
we	
  hypothesized	
  that	
  by	
  mutating	
  Thr60	
  to	
  a	
  rigid	
  amino	
  acid	
  (i.e.	
  Pro),	
  it	
  will	
  make	
  it	
  harder	
  for	
  
the	
  loop	
  to	
  close	
  back	
  and	
  form	
  monomer	
  then	
  increases	
  the	
  dimer	
  formation.	
  To	
  this	
  end	
  we	
  
mutated	
   and	
   expressed	
   T60P,	
   however	
   our	
   hypothesis	
   seems	
   not	
   to	
   be	
   right	
   and	
   it	
   also	
   only	
  
formed	
   monomer.	
   But	
   this	
   could	
   be	
   because	
   Pro	
   is	
   not	
   hydrophilic	
   and	
   since	
   this	
   residue	
   is	
  
pointed	
   toward	
   the	
   solvent	
   in	
   dimer,	
   it	
   needed	
   to	
   be	
   mutated	
   to	
   more	
   hydrophilic	
   residues	
  
instead.	
  Other	
  residues,	
  which	
  we	
  found	
  them	
  important,	
  are	
  Lys62,	
  Thr63	
  and	
  Phe66	
  because	
  
they	
   have	
   different	
   conformation	
   in	
   dimer	
   from	
   monomer.	
   It	
   should	
   be	
   considered	
   that	
   we	
   are	
  
not	
   able	
   to	
   identify	
   the	
   propensity	
   of	
   the	
   dimer	
   the	
   same	
   way	
   as	
   hCRBPII	
   (which	
   descried	
   in	
  
chapter	
  II)	
  since	
  we	
  cannot	
  separate	
  dimer	
  vs.	
  monomer	
  in	
  purification	
  steps	
  but	
  we	
  only	
  can	
  
distinguish	
  the	
  mutants	
  that	
  generate	
  only	
  one	
  of	
  the	
  species:	
  dimer	
  or	
  monomer.	
  	
  
Together,	
   this	
   data	
   will	
   probe	
   the	
   importance	
   of	
   establishing	
   the	
   phase	
   relationship	
  

110	
  

Residue Position
CRBPII Mutant
48 49 50 51 52 53 54 55 56 57 58 59 60
WT sequence
N F K T K T T S T F R N Y
Monomer
WT
O I O I O I O I L L I O I
Asymmetric for	
  
Y60WA
O group	
  
I O still	
  
I is	
  
O focusing	
  
I O oI n	
  tIhis	
  Oaim.	
  
I
required	
  
dimerization	
  Oand	
  I our	
  
Dimers
Y60WB
O I O I O I O I O I O I S
O I O I O I O I O I O I I
	
   WT
Y60L
O I O I O I O I O Residue
I O Position
I S
Symmetric KL:T51F
O I50 O51 I 52 O 53 I 54O 55I 56
O 57
I S58
CRBPII Mutant O 48I 49
	
  
Dimers
K:T51D
O
I
O
I
O
I
O
I
O
I
O
WT sequence
N F K T K T T S T IF SR
O OI OI IO O I I O O I I OO I I OL IL S I
Monomer KL:T51W
WT
E72A
O I O I O I O I O I O I I

Asymmetric Y60WA
FABP5
Dimers
Y60WB
FABP5
WT sequence
Monomer
WT-FABP5
Y60L
Symmetric WT(AZO)
Symmetric KL:T51F
Dimers
WT(AZR)
Dimers
K:T51D
Monomer
FABP9
KL:T51W
Dimer
FABP9
E72A

Possible Dimer
Possible Dimer
Possible
Dimer
Monomer
Possible
Dimer
Symmetric
Possible Dimer

Dimers

BLAST sequences
FABP5Mole
Golden
Aardvark
FABP5 sequence
Chinese
Hamster
WT-FABP5
Naked
Mole-Rat
WT(AZO)
Damara Mole-Rat

WT(AZR)

52
N
O
O
O
O
O
V
V
K
T
T

61 62 63
D V D
O I O
O I O
O I O
O I O
O I O
O59 I 60 O
ON I YO
OO I I O
O I O

64
F
I
I
I
I
I
61I
DI
OI
I

65
T
O
O
O
O
O
62O 63 64 65
VO D F T
IO O I O
O

I
I69
IT
IO
O
I
O
I
IO
IO

O
O
O
O
O
O
O
O

65F
QF
OF
IF
IF

66K
FK
IK
IT
IK

67 68 69
S C T
O I O
O I O
O I O

O
O53
OL
OI
I
O
I
O
OI
OI

I
54I
TI
O
I
O
I
O
I
OI
OI

O
55O
IO
I
O
I
O
I
O
IO
IO

I
56 I
KI
O
I
O
I
O
I
OI
OI

O
57 O 58
TOE
I O
O
I O
O
I O
O
I OO
I OO

I
I 59
IS
II
I
I
I
I
II
II

O
O60
OT
OL
O
O
O
O
OL
OO

I
61
I
IV
IL
I
I
I
I
IL
II

O
62
O
OK
OI
O
O
O
O
OI
OO

I
63I
TI
O
I
I
I
I
I
OI
II

I
64O
TO
I
O
O
O
O
O
IO
OO

O I
65 I 66 S 67
QI F IS
O I O
I S
I I O
I S
I I O
I S
O I I SO
I I I IO

52V
NV
OV
OV
OV

53
S
SL
NI
TI
T
I

I54
IT
IO
IO
I

K55
GI
QI
QI
Q

T 56 E 57S
T K E TS
TOE IS
TOE IS
T Q S

58S
ES
OS
OI
OI

59F
SF
IF
II
II

60K
KT
ML
OK
OK

N61
NV
NL
NI
N

T62
TK
TI
TO
T

M63
LT
VO
VI
V

O

I

O

I

I

O

I

I
I
I
I
I

64 S
TT
IS
OS
OS

O
O68
OC
OI
I
O
I
O
OI
OI

I
I
I
I
I
I
I
I

O
O
O
O
O
O
O
O

I: Inside; O: Outside; S: Neither Inside or Outside, but Solvent exposed; L: loop

Monomer
FABP9
O I O I O I O I L L I O I O I O I O
DimerIII-­‐11:	
  FABP9
O in	
  I iLPBP	
  
O fI amily.	
  
O I O I O I O I O I I O I O
Figure	
  
Phase	
  relationship	
  
BLAST sequences
Possible
Golden Mole
V V S I K T E S S F K N T M I S F K
III-­‐4	
  EDimer
xperimental	
  
Possible Dimer Aardvark
V V S I G T E S S F K N T L I T F K
III-­‐4-­‐1	
  
Material	
  
nd	
  Method:	
  
Possible
Dimer
ChineseaHamster
K SVite-­‐Directed	
  
N I Q M
T utagenesis	
  
E S S F M N T V I S F K
Possible Dimer Naked Mole-Rat
T V T I Q T E S I I K N T V I S F T
The	
  
hFABP5	
  
and	
  
hFABP4	
  
plasmids	
  
Possible Dimer Damara Mole-Rat
T V described	
  
T I Q were	
  
T Q ordered	
  
S I from	
  
I KIDTÂŽ.	
  
N Then	
  
T V we	
  I cloned	
  
S F them	
  
K
and	
   HindIII	
  
I: into	
  
Inside; pET28a	
  
O: Outside; S:(Novagen)	
  
Neither Inside between	
  
or Outside, butBamHI	
  
Solvent exposed;
L: loop cut	
   sites	
   and	
   it	
   yielded	
   N-­‐terminal	
   his-­‐
tagged	
  (cleavable	
  by	
  thrombin).	
  We	
  used	
  these	
  constructs	
  for	
  mutagenesis	
  as	
  well.	
  	
  (Detailed	
  
protocol	
  explained	
  in	
  section	
  III-­‐7).	
  
	
  
	
  

111	
  

Synthesized	
   amino	
   acid	
   sequence	
   of	
   hFABP5	
   (N-­‐terminal	
   Histagged-­‐thrombin):	
  
HHHHHHSSGLVPRGSHMGSMGSHMATVQQLEGRWRLVDSKGFDEYMKELGVGIALRKMGAMAKPDCII
TCDGKNLTIKTESTVKTTQFSCTLGEKFEETTADGRKTQTVCNFTDGALVQHQEWDGKESTITRKLKDGKLVV
ECVMNNVTCTRIYEKVE	
  
Primers	
  	
  
T64A-­‐hFABP5	
  	
  
Forward:	
  5ʼ-­‐	
  CGTGAAGACGGCGCAGTTTTCAT-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  GTCGATTCAGTTTTAATAGTTAAGGTTCTTA-­‐3ʼ	
  
	
  
T60P-­‐hFABP5	
  
Forward:	
  5ʼ-­‐	
  AACTGAATCGCCCGTGAAGACGA	
  -­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TTAATAGTTAAGTTCTTACCATCGC-­‐3ʼ	
  
	
  
	
  
T64E-­‐hFABP5	
  
Forward:	
  5’-­‐	
  CGTGAAGACGGAGCAGTTTTCATG-­‐3’	
  
Reverse:	
  5’-­‐	
  GTCGATTCAGTTTTAATAGTTAAG-­‐3’	
  
	
  
T64L-­‐hFABP5	
  
Forward:	
  5’-­‐	
  CGTGGAGACGCTGCAGTTTTCATG-­‐	
  3’	
  
Reverse:	
  5’-­‐	
  GTCGATTCAGTTTTAATAGTTAAG-­‐	
  3’	
  
	
  
Q65W-­‐hFABP5	
  
Forward:	
  5’-­‐	
  GAAGACGACGTGGTTTTCATGCAC-­‐3’	
  	
  
Reverse:	
  5’-­‐ACGGTCGATTCAGTTTTAATAG	
  -­‐3’	
  	
  
	
  
Q65M-­‐hFABP5	
  
Forward:	
  5’-­‐GAAGACGATGTTTTCATGCAC-­‐3’	
  	
  
Reverse:	
  5’-­‐ACGGTCGATTCAGTTTTAATAG-­‐3’	
  	
  
	
  
Q65W-­‐hFABP5	
  
Forward:	
  5’-­‐	
  GAAGACGACGTGGTTTTCATGCAC-­‐3’	
  	
  
Reverse:	
  5’-­‐ACGGTCGATTCAGTTTTAATAG	
  -­‐3’	
  	
  
	
  
Q65M-­‐hFABP5	
  
Forward:	
  5’-­‐GAAGACGATGTTTTCATGCAC-­‐3’	
  	
  
Reverse:	
  5’-­‐ACGGTCGATTCAGTTTTAATAG-­‐3’	
  	
  
	
  

112	
  

III-­‐4-­‐2	
  Protein	
  Expression	
  and	
  Purification	
  of	
  hFABP4	
  and	
  hFABP5	
  
Proteins	
   were	
   expressed	
   in	
   Escherichia	
   coli	
   BL21	
   (DE3)	
   cells	
   using	
   the	
   T7	
   expression	
   system.	
  
Cells	
   were	
   grown	
   in	
   LB	
   media	
   and	
   were	
   induced	
   at	
   OD600	
   of	
   0.7,	
   by	
   adding	
   IPTG	
   to	
   a	
   final	
  
concentration	
   of	
   0.4	
   mM.	
   After	
   20	
   h	
   incubation	
   at	
   20°C,	
   the	
   cells	
   were	
   collected	
   by	
  
centrifugation	
  at	
  5000rpm	
  at	
  4°C	
  for	
  30	
  min.	
  Cells	
  were	
  lysed	
  in	
  30	
  ml	
  ice-­‐cold	
  column	
  buffer:	
  
20mM	
  Tris	
  pH	
  8.5	
  and	
  250	
  mM	
  NaCl	
  (20mM	
  Tris	
  pH	
  7.4,	
  500mM	
  NaCl	
  and	
  5mM	
  imidazole	
  in	
  
the	
  case	
  of	
  hFABP4),	
  by	
  sonication	
  on	
  ice.	
  Then	
  followed	
  by	
  30	
  min	
  centrifugation	
  at	
  10,000rpm	
  
at	
  4°C.	
  The	
  supernatant	
  was	
  loaded	
  onto	
  a	
  Ni–NTA	
  column	
  (GE	
  healthcare)	
  and	
  batched	
  bind	
  
for	
   one	
   hour	
   at	
   4°C.	
   The	
   column	
   was	
   washed	
   with	
   10CV	
   buffer	
   containing	
   20	
   mM	
   imidazole.	
  
The	
   proteins	
   were	
   then	
   eluted	
   with	
   column	
   buffer	
   containing	
   200	
   mM	
   imidazole.	
   In	
   case	
   of	
  
delipidation,	
   protein	
   were	
   added	
   to	
   Lipidex-­‐5000	
   resin	
   and	
   incubated	
   for	
   1	
   hour	
   at	
   37°C	
   and	
  
collected	
   the	
   flow	
   through.	
   (sigma)	
   	
   The	
   affinity-­‐purified	
   proteins	
   were	
   incubated	
   with	
  
thrombin	
  (GE	
  Healthcare	
  Life	
  Sciences)	
  at	
  10	
  units	
  per	
  milligram	
  of	
  protein	
  at	
  4°C	
  overnight	
  in	
  
order	
  to	
  remove	
  the	
  6	
  His	
  tagged.	
  The	
  cleaved	
  protein	
  solutions	
  were	
  loaded	
  onto	
  an	
  Ni–NTA	
  
columns	
  and	
  the	
  unbound	
  fractions	
  of	
  cleaved	
  materials	
  collected	
  and	
  then	
  concentrated.	
  The	
  
affinity-­‐purified	
   samples	
   were	
   concentrated	
   to	
   about	
   2mL	
   and	
   loaded	
   onto	
   our	
   SEC	
   column	
  
(Superdex	
   S75	
   16/600	
   HiLoad	
   column,	
   GE	
   healthcare)	
   equilibrated	
   with	
   1CV	
   PBS	
   pH	
   8.5.	
   The	
  
peak	
   fractions	
   were	
   collected.	
   (In	
   the	
   case	
   of	
   hFABP5,	
   only	
   few	
   times	
   delipidated	
   by	
   incubation	
  
with	
   Lipidex-­‐	
   5000,	
   Sigma	
   at	
   37°	
   C	
   for	
   1	
   h	
   with	
   occasional	
   mixing	
   followed	
   by	
   second	
   run	
   of	
  
SEC).	
   Almost	
   all	
   mentioned	
   (except	
   T64L,	
   which	
   didn’t	
   express	
   well)	
   mutations,	
   purified	
  
proteins	
  were	
  concentrated	
  to	
  approximately	
  10mg	
  /mL	
  for	
  crystallization	
  purpose.	
  

113	
  

III-­‐4-­‐3	
  Crystallization,	
  data	
  collection	
  and	
  refinement	
  
The	
  concentrated	
  un-­‐cleaved	
  human	
  FABP5	
  apo	
  was	
  crystallized	
  in	
  25%	
  PEG	
  3350,	
  100	
  mM	
  bis-­‐
tris	
  pH	
  6.0.	
  For	
  details	
  on	
  data	
  collection,	
  processing	
  the	
  data	
  and	
  structure	
  refinement	
  please	
  
refer	
  to	
  section	
  II-­‐7-­‐8	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

114	
  

!
!
Table	
  
! III-­‐1:	
  Crystallographic	
  data	
  of	
  hFABP5	
  (pseudo	
  monomer).	
  
	
  
WT$hFABP5! ! 	
  
!
Space!group!
! P21!21!21! 	
   !
	
  
a!(Å)!
! 40.89!
!
	
  
b!(Å)!
! 55.10! 	
  	
  
!
c!(Å)!
! 126.36! 	
  
!
	
  
ι!(°)!
! 90!
!
	
  
β!(°)!
! 90! 	
  
!
δ!(°)!
! 90! 	
  
!
	
  
Molecules!per!
! 	
  
!
	
  
asymmetric!unit!
2!
	
  
Total!reflection!
! 1524152! 	
  
!
	
  
Unique!reflection!
! 4421!
!
!a 	
  
Completeness!(%)!
! 99.41!(92) ! 	
   !
	
  
Average!I/σ!
! 30! 	
  
!
!
	
   !
	
  
! 34.329$
!
	
  
Resolution!(Å)!(last!
1.789!(1.83$ 	
  
shell)!
1.78)!a! 	
   	
  
!
Rwork/Rfree!(%)!
25/28.2! 	
  
	
  
Root$mean$square! !
!
!
	
  
deviation!from!ideal! 	
  
values!
!
Bond!length!(Å)!
0.009!
!
! 1.394!
!
Bond!angle!(°)!
! 27.15!
!
Average!B!factor!
!

!

!
!

!

!
!

!

115	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
REFERENCES	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

116	
  

REFERENCES	
  
	
  
	
  
	
  
1.	
  
	
  
2.	
  
	
  
3.	
  

	
  
4.	
  
	
  
5.	
  
	
  
6.	
  

	
  
7.	
  
	
  
8.	
  
	
  
9.	
  
	
  
10.	
  
	
  
11.	
  

Ong,	
  D.	
  E.	
  (1984)	
  A	
  NOVEL	
  RETINOL-­‐BINDING	
  PROTEIN	
  FROM	
  RAT	
  -­‐	
  PURIFICATION	
  AND	
  
PARTIAL	
  CHARACTERIZATION,	
  Journal	
  of	
  Biological	
  Chemistry	
  259,	
  1476-­‐1482.	
  
Harder,	
  M.	
  E.,	
  Deinzer,	
  M.	
  L.,	
  Leid,	
  M.	
  E.,	
  and	
  Schimerlik,	
  M.	
  I.	
  (2004)	
  Global	
  analysis	
  of	
  
three-­‐state	
  protein	
  unfolding	
  data,	
  Protein	
  Science	
  13,	
  2207-­‐2222.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  crystallographica.	
  Section	
  
D,	
  Biological	
  crystallography	
  70,	
  3226-­‐3232.	
  
Veerkamp,	
   J.	
   H.,	
   Vankuppevelt,	
   T.,	
   Maatman,	
   R.,	
   and	
   Prinsen,	
   C.	
   F.	
   M.	
   (1993)	
  
STRUCTURAL	
   AND	
   FUNCTIONAL-­‐ASPECTS	
   OF	
   CYTOSOLIC	
   FATTY-­‐ACID-­‐BINDING	
  
PROTEINS,	
  Prostaglandins	
  Leukotrienes	
  and	
  Essential	
  Fatty	
  Acids	
  49,	
  887-­‐906.	
  
Banaszak,	
   L.,	
   Winter,	
   N.,	
   Xu,	
   Z.	
   H.,	
   Bernlohr,	
   D.	
   A.,	
   Cowan,	
   S.,	
   and	
   Jones,	
   T.	
   A.	
   (1994)	
  
LIPID-­‐BINDING	
   PROTEINS	
   -­‐	
   A	
   FAMILY	
   OF	
   FATTY-­‐ACID	
   AND	
   RETINOID	
   TRANSPORT	
  
PROTEINS,	
  Advances	
  in	
  Protein	
  Chemistry,	
  Vol	
  45	
  45,	
  89-­‐151.	
  
Vaezeslami,	
   S.,	
   Mathes,	
   E.,	
   Vasilelou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2006)	
   The	
  
structure	
   of	
   apo-­‐wild-­‐type	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   at	
   1.4	
   A	
   and	
   its	
  
relationship	
   to	
   ligand	
   binding	
   and	
   nuclear	
   translocation,	
   Journal	
   of	
   Molecular	
   Biology	
  
363,	
  687-­‐701.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins-­‐-­‐insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  lipid	
  research	
  43,	
  328-­‐349.	
  
Lucke,	
   C.,	
   Huang,	
   S.,	
   Rademacher,	
   M.,	
   and	
   Ruterjans,	
   H.	
   (2002)	
   New	
   insights	
   into	
  
intracellular	
   lipid	
   binding	
   proteins:	
   The	
   role	
   of	
   buried	
   water,	
   Protein	
   science	
   :	
   a	
  
publication	
  of	
  the	
  Protein	
  Society	
  11,	
  2382-­‐2392.	
  
Schaap,	
   F.	
   G.,	
   van	
   der	
   Vusse,	
   G.	
   J.,	
   and	
   Glatz,	
   J.	
   F.	
   (2002)	
   Evolution	
   of	
   the	
   family	
   of	
  
intracellular	
   lipid	
   binding	
   proteins	
   in	
   vertebrates,	
   Molecular	
   and	
   cellular	
   biochemistry	
  
239,	
  69-­‐77.	
  
Smathers,	
   R.	
   L.,	
   and	
   Petersen,	
   D.	
   R.	
   (2011)	
   The	
   human	
   fatty	
   acid-­‐binding	
   protein	
   family:	
  
evolutionary	
  divergences	
  and	
  functions,	
  Human	
  genomics	
  5,	
  170-­‐191.	
  
Franzoni,	
   L.,	
   Lucke,	
   C.,	
   Perez,	
   C.,	
   Cavazzini,	
   D.,	
   Rademacher,	
   M.,	
   Ludwig,	
   C.,	
   Spisni,	
   A.,	
  
Rossi,	
  G.	
  L.,	
  and	
  Ruterjans,	
  H.	
  (2002)	
  Structure	
  and	
  backbone	
  dynamics	
  of	
  apo-­‐	
  and	
  holo-­‐
117	
  

	
  
12.	
  

	
  
13.	
  

	
  
14.	
  

	
  
15.	
  
	
  
16.	
  
	
  
17.	
  

	
  
18.	
  
	
  
19.	
  
	
  
20.	
  
	
  
21.	
  

cellular	
   retinol-­‐binding	
   protein	
   in	
   solution,	
   Journal	
   of	
   Biological	
   Chemistry	
   277,	
   21983-­‐
21997.	
  
Crow,	
   J.	
   A.,	
   and	
   Ong,	
   D.	
   E.	
   (1985)	
   CELL-­‐SPECIFIC	
   IMMUNOHISTOCHEMICAL	
  
LOCALIZATION	
   OF	
   A	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   (TYPE-­‐2)	
   IN	
   THE	
   SMALL-­‐
INTESTINE	
  OF	
  RAT,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  
of	
  America	
  82,	
  4707-­‐4711.	
  
Vogel,	
  S.,	
  Mendelsohn,	
  C.	
  L.,	
  Mertz,	
  J.	
  R.,	
  Piantedosi,	
  R.,	
  Waldburger,	
  C.,	
  Gottesman,	
  M.	
  
E.,	
  and	
  Blaner,	
  W.	
  S.	
  (2001)	
  Characterization	
  of	
  a	
  new	
  member	
  of	
  the	
  fatty	
  acid-­‐binding	
  
protein	
   family	
   that	
   binds	
   all-­‐trans-­‐retinol,	
   Journal	
   of	
   Biological	
   Chemistry	
   276,	
   1353-­‐
1360.	
  
Folli,	
   C.,	
   Calderone,	
   V.,	
   Ottonello,	
   S.,	
   Bolchi,	
   A.,	
   Zanotti,	
   G.,	
   Stoppini,	
   M.,	
   and	
   Berni,	
   R.	
  
(2001)	
   Identification,	
   retinoid	
   binding,	
   and	
   x-­‐ray	
   analysis	
   of	
   a	
   human	
   retinol-­‐binding	
  
protein,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  98,	
  3710-­‐3715.	
  
Li,	
  E.,	
  Demmer,	
  L.	
  A.,	
  Sweetser,	
  D.	
  A.,	
  Ong,	
  D.	
  E.,	
  and	
  Gordon,	
  J.	
  I.	
  (1986)	
  ISOLATION	
  OF	
  A	
  
CLONED	
   CDNA-­‐ENCODING	
   AN	
   ABUNDANT	
   INTESTINAL	
   RETINOL	
   BINDING-­‐PROTEIN,	
  
Clinical	
  Research	
  34,	
  A685-­‐A685.	
  
Li,	
   E.,	
   and	
   Norris,	
   A.	
   W.	
   (1996)	
   Structure/function	
   of	
   cytoplasmic	
   vitamin	
   A-­‐binding	
  
proteins,	
  Annual	
  Review	
  of	
  Nutrition	
  16,	
  205-­‐234.	
  
Li,	
   E.,	
   Qian,	
   S.	
   J.,	
   Winter,	
   N.	
   S.,	
   Davignon,	
   A.,	
   Levin,	
   M.	
   S.,	
   and	
   Gordon,	
   J.	
   I.	
   (1991)	
  
FLUORINE	
   NUCLEAR-­‐MAGNETIC-­‐RESONANCE	
   ANALYSIS	
   OF	
   THE	
   LIGAND-­‐BINDING	
  
PROPERTIES	
   OF	
   2	
   HOMOLOGOUS	
   RAT	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEINS	
  
EXPRESSED	
  IN	
  ESCHERICHIA-­‐COLI,	
  Journal	
  of	
  Biological	
  Chemistry	
  266,	
  3622-­‐3629.	
  
Calderone,	
   V.,	
   Folli,	
   C.,	
   Marchesani,	
   A.,	
   Berni,	
   R.,	
   and	
   Zanotti,	
   G.	
   (2002)	
   Identification	
  
and	
   structural	
   analysis	
   of	
   a	
   zebrafish	
   apo	
   and	
   holo	
   cellular	
   retinol-­‐binding	
   protein,	
  
Journal	
  of	
  Molecular	
  Biology	
  321,	
  527-­‐535.	
  
Tarter,	
  M.,	
  Capaldi,	
  S.,	
  Carrizo,	
  M.	
  E.,	
  Ambrosi,	
  E.,	
  Perduca,	
  M.,	
  and	
  Monaco,	
  H.	
  L.	
  (2008)	
  
Crystal	
   structure	
   of	
   human	
   cellular	
   retinol-­‐binding	
   protein	
   11	
   to	
   1.2	
   angstrom	
  
resolution,	
  Proteins-­‐Structure	
  Function	
  and	
  Bioinformatics	
  70,	
  1626-­‐1630.	
  
Winter,	
   N.	
   S.,	
   Bratt,	
   J.	
   M.,	
   and	
   Banaszak,	
   L.	
   J.	
   (1993)	
   CRYSTAL-­‐STRUCTURES	
   OF	
   HOLO	
  
AND	
   APO-­‐CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN-­‐II,	
   Journal	
   of	
   Molecular	
   Biology	
   230,	
  
1247-­‐1259.	
  
Kane,	
  M.	
  A.,	
  Bright,	
  F.	
  V.,	
  and	
  Napoli,	
  J.	
  L.	
  (2011)	
  Binding	
  affinities	
  of	
  CRBPI	
  and	
  CRBPII	
  
for	
  9-­‐cis-­‐retinoids,	
  Biochimica	
  Et	
  Biophysica	
  Acta-­‐General	
  Subjects	
  1810,	
  514-­‐518.	
  
118	
  

22.	
  
	
  
23.	
  

	
  
24.	
  
	
  
25.	
  
	
  
26.	
  
	
  
27.	
  

	
  
28.	
  
	
  
29.	
  
	
  
30.	
  
	
  
31.	
  
	
  
32.	
  

Malpeli,	
   G.,	
   Stoppini,	
   M.,	
   Zapponi,	
   M.	
   C.,	
   Folli,	
   C.,	
   and	
   Berni,	
   R.	
   (1995)	
   INTERACTIONS	
  
WITH	
   RETINOL	
   AND	
   RETINOIDS	
   OF	
   BOVINE	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN,	
  
European	
  Journal	
  of	
  Biochemistry	
  229,	
  486-­‐493.	
  
McBee,	
   J.	
   K.,	
   Kuksa,	
   V.,	
   Alvarez,	
   R.,	
   de	
   Lera,	
   A.	
   R.,	
   Prezhdo,	
   O.,	
   Haeseleer,	
   F.,	
   Sokal,	
   I.,	
  
and	
   Palczewski,	
   K.	
   (2000)	
   Isomerization	
   of	
   all-­‐trans-­‐retinol	
   to	
   cis-­‐retinols	
   in	
   bovine	
  
retinal	
   pigment	
   epithelial	
   cells:	
   Dependence	
   on	
   the	
   specificity	
   of	
   retinoid-­‐binding	
  
proteins,	
  Biochemistry	
  39,	
  11370-­‐11380.	
  
Borshchevskiy,	
  V.	
  I.,	
  Round,	
  E.	
  S.,	
  Popov,	
  A.	
  N.,	
  Bueldt,	
  G.,	
  and	
  Gordeliy,	
  V.	
  I.	
  (2011)	
  X-­‐
ray-­‐Radiation-­‐Induced	
   Changes	
   in	
   Bacteriorhodopsin	
   Structure,	
   Journal	
   of	
   Molecular	
  
Biology	
  409,	
  813-­‐825.	
  
Matsui,	
  Y.,	
  Sakai,	
  K.,	
  Murakami,	
  M.,	
  Shiro,	
  Y.,	
  Adachi,	
  S.,	
  Okumura,	
  H.,	
  and	
  Kouyama,	
  T.	
  
(2002)	
  Specific	
  damage	
  induced	
  by	
  X-­‐ray	
  radiation	
  and	
  structural	
  changes	
  in	
  the	
  primary	
  
photoreaction	
  of	
  bacteriorhodopsin,	
  Journal	
  of	
  Molecular	
  Biology	
  324,	
  469-­‐481.	
  
Schreckenbach,	
   T.,	
   Walckhoff,	
   B.,	
   and	
   Oesterhelt,	
   D.	
   (1977)	
   STUDIES	
   ON	
   RETINAL-­‐
PROTEIN	
  INTERACTION	
  IN	
  BACTERIORHODOPSIN,	
  European	
  Journal	
  of	
  Biochemistry	
  76,	
  
499-­‐511.	
  
Cowan,	
   S.	
   W.,	
   Newcomer,	
   M.	
   E.,	
   and	
   Jones,	
   T.	
   A.	
   (1993)	
   CRYSTALLOGRAPHIC	
   STUDIES	
  
ON	
   A	
   FAMILY	
   OF	
   CELLULAR	
   LIPOPHILIC	
   TRANSPORT	
   PROTEINS	
   -­‐	
   REFINEMENT	
   OF	
   P2	
  
MYELIN	
   PROTEIN	
   AND	
   THE	
   STRUCTURE	
   DETERMINATION	
   AND	
   REFINEMENT	
   OF	
  
CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   IN	
   COMPLEX	
   WITH	
   ALL-­‐TRANS-­‐RETINOL,	
   Journal	
  
of	
  Molecular	
  Biology	
  230,	
  1225-­‐1246.	
  
Vasileiou,	
   C.,	
   Vaezeslami,	
   S.,	
   Crist,	
   R.	
   M.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2007)	
   Protein	
   design:	
   reengineering	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   into	
   a	
  
rhodopsin	
  protein	
  mimic,	
  J	
  Am	
  Chem	
  Soc	
  129,	
  6140-­‐6148.	
  
Otwinowski,	
   Z.,	
   and	
   Minor,	
   W.	
   (1997)	
   Processing	
   of	
   X-­‐ray	
   diffraction	
   data	
   collected	
   in	
  
oscillation	
  mode,	
  Macromolecular	
  Crystallography,	
  Pt	
  A	
  276,	
  307-­‐326.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (1997)	
   MOLREP:	
   an	
   automated	
   program	
   for	
   molecular	
  
replacement,	
  Journal	
  of	
  Applied	
  Crystallography	
  30,	
  1022-­‐1025.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (2010)	
   Molecular	
   replacement	
   with	
   MOLREP,	
   Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  22-­‐25.	
  
Murshudov,	
   G.	
   N.,	
   Skubak,	
   P.,	
   Lebedev,	
   A.	
   A.,	
   Pannu,	
   N.	
   S.,	
   Steiner,	
   R.	
   A.,	
   Nicholls,	
   R.	
   A.,	
  
Winn,	
   M.	
   D.,	
   Long,	
   F.,	
   and	
   Vagin,	
   A.	
   A.	
   (2011)	
   REFMAC5	
   for	
   the	
   refinement	
   of	
  
macromolecular	
   crystal	
   structures,	
   Acta	
   Crystallographica	
   Section	
   D-­‐Biological	
  
Crystallography	
  67,	
  355-­‐367.	
  
119	
  

33.	
  
	
  
34.	
  
	
  
35.	
  
	
  
36.	
  
	
  
37.	
  
	
  
38.	
  
	
  
39.	
  
	
  
40.	
  
	
  
41.	
  
	
  
42.	
  
	
  
43.	
  
	
  
44.	
  
	
  
45.	
  

Emsley,	
   P.,	
   Lohkamp,	
   B.,	
   Scott,	
   W.	
   G.,	
   and	
   Cowtan,	
   K.	
   (2010)	
   Features	
   and	
   development	
  
of	
  Coot,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  486-­‐501.	
  
Bennett,	
   M.	
   J.,	
   Schlunegger,	
   M.	
   P.,	
   and	
   Eisenberg,	
   D.	
   (1995)	
   3D	
   domain	
   swapping	
   -­‐	
   A	
  
mechanism	
  for	
  oligomer	
  assembly,	
  Protein	
  Science	
  4,	
  2455-­‐2468.	
  
Crestfield,	
   A.	
   M.,	
   Stein,	
   W.	
   H.,	
   and	
   Moore,	
   S.	
   (1962)	
   On	
   the	
   aggregation	
   of	
   bovine	
  
pancreatic	
  ribonuclease,	
  Archives	
  of	
  biochemistry	
  and	
  biophysics	
  Suppl	
  1,	
  217-­‐222.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   As	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  Science	
  11,	
  1285-­‐1299.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   Structures	
   of	
   the	
   two	
   3D	
  
domain-­‐swapped	
  RNase	
  A	
  trimers,	
  Protein	
  Science	
  11,	
  371-­‐380.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   S.	
   (2001)	
   Crystal	
   structure	
   of	
   an	
  
Rnase	
  A	
  dimer	
  displaying	
  a	
  new	
  type	
  of	
  3D	
  domain	
  swapping,	
  Nucleic	
  Acid	
  Database.	
  
Liu,	
  Y.	
  S.,	
  Hart,	
  P.	
  J.,	
  Schlunegger,	
  M.	
  P.,	
  and	
  Eisenberg,	
  D.	
  (1998)	
  The	
  crystal	
  structure	
  of	
  
a	
   3D	
   domain-­‐swapped	
   dimer	
   of	
   RNase	
   A	
   at	
   a	
   2.1-­‐angstrom	
   resolution,	
   Proceedings	
   of	
  
the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  95,	
  3437-­‐3442.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   as	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  11,	
  1285-­‐1299.	
  
Koharudin,	
   L.	
   M.,	
   Liu,	
   L.,	
   and	
   Gronenborn,	
   A.	
   M.	
   (2013)	
   Different	
   3D	
   domain-­‐swapped	
  
oligomeric	
   cyanovirin-­‐N	
   structures	
   suggest	
   trapped	
   folding	
   intermediates,	
   Proc	
   Natl	
  
Acad	
  Sci	
  U	
  S	
  A	
  110,	
  7702-­‐7707.	
  
Assar,	
  Z.,	
  Nossoni,	
  Z.,	
  Wang,	
  W.,	
  Santos,	
  E.	
  M.,	
  Kramer,	
  K.,	
  McCornack,	
  C.,	
  Vasileiou,	
  C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   Domain-­‐Swapped	
   Dimers	
   of	
   Intracellular	
   Lipid-­‐
Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure	
  24,	
  1590-­‐1598.	
  
Bax,	
  B.,	
  Lapatto,	
  R.,	
  Nalini,	
  V.,	
  Driessen,	
  H.,	
  Lindley,	
  P.	
  F.,	
  Mahadevan,	
  D.,	
  Blundell,	
  T.	
  L.,	
  
and	
   Slingsby,	
   C.	
   (1990)	
   X-­‐ray	
   analysis	
   of	
   beta-­‐B2-­‐crystallin	
   and	
   evolution	
   of	
   oligomeric	
  
lens	
  proteins,	
  Nature	
  347,	
  776-­‐780.	
  
Hakansson,	
  M.,	
  Svensson,	
  A.,	
  Fast,	
  J.,	
  and	
  Linse,	
  S.	
  (2001)	
  An	
  extended	
  hydrophobic	
  core	
  
induces	
  EF-­‐hand	
  swapping,	
  Protein	
  Science	
  10,	
  927-­‐933.	
  
Rousseau,	
   F.,	
   Schymkowitz,	
   J.,	
   and	
   Itzhaki,	
   L.	
   S.	
   (2012)	
   Implications	
   of	
   3D	
   domain	
  
swapping	
   for	
   protein	
   folding,	
   misfolding	
   and	
   function,	
   In	
   Protein	
   Dimerization	
   and	
  
Oligomerization	
  in	
  Biology	
  (Mattews,	
  J.	
  M.,	
  Ed.),	
  pp	
  137-­‐152.	
  

	
  
120	
  

46.	
  

	
  
47.	
  
	
  
48.	
  
	
  
49.	
  
	
  
50.	
  
	
  
51.	
  
	
  
52.	
  
53.	
  
	
  
54.	
  
	
  
55.	
  
	
  
56.	
  
	
  

Yang,	
   F.,	
   Bewley,	
   C.	
   A.,	
   Louis,	
   J.	
   M.,	
   Gustafson,	
   K.	
   R.,	
   Boyd,	
   M.	
   R.,	
   Gronenborn,	
   A.	
   M.,	
  
Clore,	
  G.	
  M.,	
  and	
  Wlodawer,	
  A.	
  (1999)	
  Crystal	
  structure	
  of	
  cyanovirin-­‐N,	
  a	
  potent	
  HIV-­‐
inactivating	
  protein,	
  shows	
  unexpected	
  domain	
  swapping,	
  Journal	
  of	
  Molecular	
  Biology	
  
288,	
  403-­‐412.	
  
Szilagyi,	
   A.,	
   Zhang,	
   Y.,	
   and	
   Zavodszky,	
   P.	
   (2012)	
   Intra-­‐Chain	
   3D	
   Segment	
   Swapping	
  
Spawns	
  the	
  Evolution	
  of	
  New	
  Multidomain	
  Protein	
  Architectures,	
  Journal	
  of	
  Molecular	
  
Biology	
  415,	
  221-­‐235.	
  
Stroud,	
   J.	
   C.,	
   Liu,	
   C.,	
   Teng,	
   P.	
   K.,	
   and	
   Eisenberg,	
   D.	
   (2012)	
   Toxic	
   fibrillar	
   oligomers	
   of	
  
amyloid-­‐beta	
   have	
   cross-­‐beta	
   structure,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  109,	
  7717-­‐7722.	
  
Cleary,	
   J.	
   P.,	
   Walsh,	
   D.	
   M.,	
   Hofmeister,	
   J.	
   J.,	
   Shankar,	
   G.	
   M.,	
   Kuskowski,	
   M.	
   A.,	
   Selkoe,	
   D.	
  
J.,	
   and	
   Ashe,	
   K.	
   H.	
   (2005)	
   Natural	
   oligomers	
   of	
   the	
   amyloid-­‐protein	
   specifically	
   disrupt	
  
cognitive	
  function,	
  Nature	
  Neuroscience	
  8,	
  79-­‐84.	
  
Fandrich,	
   M.,	
   Fletcher,	
   M.	
   A.,	
   and	
   Dobson,	
   C.	
   M.	
   (2001)	
   Amyloid	
   fibrils	
   from	
   muscle	
  
myoglobin	
   -­‐	
   Even	
   an	
   ordinary	
   globular	
   protein	
   can	
   assume	
   a	
   rogue	
   guise	
   if	
   conditions	
  
are	
  right,	
  Nature	
  410,	
  165-­‐166.	
  
Gronenborn,	
   A.	
   M.	
   (2009)	
   Protein	
   acrobatics	
   in	
   pairs	
   -­‐	
   dimerization	
   via	
   domain	
  
swapping,	
  Current	
  Opinion	
  in	
  Structural	
  Biology	
  19,	
  39-­‐49.	
  
Schlunegger,	
  M.	
  P.,	
  Bennett,	
  M.	
  J.,	
  and	
  Eisenberg,	
  D.	
  (1997)	
  Oligomer	
  formation	
  by	
  3D	
  
domain	
   swapping:	
   A	
   model	
   for	
   protein	
   assembly	
   and	
   misassembly,	
   In	
   Advances	
   in	
  
Protein	
  Chemistry,	
  Vol	
  50:	
  Protein	
  Misassembly	
  (Wetzel,	
  R.,	
  Ed.),	
  pp	
  61-­‐122.	
  
Selkoe,	
  D.	
  J.	
  (2004)	
  Folding	
  proteins	
  in	
  fatal	
  ways	
  (vol	
  426,	
  pg	
  900,	
  2003),	
  Nature	
  428,	
  
445-­‐445.	
  
Shtilerman,	
  M.	
  D.,	
  Ding,	
  T.	
  T.,	
  and	
  Lansbury,	
  P.	
  T.	
  (2002)	
  Molecular	
  crowding	
  accelerates	
  
fibrillization	
   of	
   alpha-­‐synuclein:	
   Could	
   an	
   increase	
   in	
   the	
   cytoplasmic	
   protein	
  
concentration	
  induce	
  Parkinson's	
  disease?,	
  Biochemistry	
  41,	
  3855-­‐3860.	
  
Sunde,	
  M.,	
  and	
  Blake,	
  C.	
  (1997)	
  The	
  structure	
  of	
  amyloid	
  fibrils	
  by	
  electron	
  microscopy	
  
and	
   X-­‐ray	
   diffraction,	
   In	
   Advances	
   in	
   Protein	
   Chemistry,	
   Vol	
   50:	
   Protein	
   Misassembly	
  
(Wetzel,	
  R.,	
  Ed.),	
  pp	
  123-­‐159.	
  
Bennett,	
  M.	
  J.,	
  Choe,	
  S.,	
  and	
  Eisenberg,	
  D.	
  (1994)	
  Refined	
  structure	
  of	
  dimeric	
  diphtheria	
  
toxin	
  at	
  2.0	
  A	
  resolution,	
  Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  3,	
  1444-­‐
1463.	
  

121	
  

57.	
  
	
  
58.	
  
	
  
59.	
  
	
  
60.	
  
	
  
61.	
  
	
  
62.	
  
	
  
63.	
  
	
  
64.	
  
65.	
  
	
  
66.	
  
	
  
67.	
  
	
  
68.	
  
	
  
69.	
  
	
  

O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
  environment,	
  Curr	
  Opin	
  Struct	
  Biol	
  11,	
  114-­‐119.	
  
Bennett,	
   M.	
   J.,	
   Choe,	
   S.,	
   and	
   Eisenberg,	
   D.	
   (1994)	
   Rlefined	
   strucutre	
   of	
   dimeric	
  
diphtheria-­‐toxin	
  at	
  2.0-­‐angstrom	
  resolution,	
  Protein	
  Science	
  3,	
  1444-­‐1463.	
  
Carroll,	
   S.	
   F.,	
   Barbieri,	
   J.	
   T.,	
   and	
   Collier,	
   R.	
   J.	
   (1986)	
   Dimeric	
   form	
   of	
   diphtheria-­‐toxin-­‐
purification	
  and	
  characterization,	
  Biochemistry	
  25,	
  2425-­‐2430.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
   environment	
   (vol	
   11,	
   pg	
   114,	
   2001),	
   Current	
   Opinion	
   in	
   Structural	
   Biology	
  
11,	
  500-­‐500.	
  
Minton,	
   A.	
   P.	
   (2000)	
   Implications	
   of	
   macromolecular	
   crowding	
   for	
   protein	
   assembly,	
  
Current	
  Opinion	
  in	
  Structural	
  Biology	
  10,	
  34-­‐39.	
  
O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   J.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nature	
  chemical	
  biology	
  10,	
  884-­‐891.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2004)	
  Monitoring	
  protein	
  stability	
  and	
  aggregation	
  in	
  
vivo	
  by	
  real-­‐time	
  fluorescent	
  labeling,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  101,	
  523-­‐528.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2006)	
  Inhibition	
  of	
  protein	
  aggregation	
  in	
  vitro	
  and	
  in	
  
vivo	
  by	
  a	
  natural	
  osmoprotectant,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  103,	
  13357-­‐13361.	
  
Ropson,	
  I.	
  J.,	
  Boyer,	
  J.	
  A.,	
  Schaeffer,	
  B.	
  A.,	
  and	
  Dalessio,	
  P.	
  M.	
  (2009)	
  Comparison	
  of	
  the	
  
folding	
   mechanism	
   of	
   highly	
   homologous	
   proteins	
   in	
   the	
   lipid-­‐binding	
   protein	
   family,	
  
Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  75,	
  799-­‐806.	
  
Burns,	
   L.	
   L.,	
   and	
   Ropson,	
   I.	
   J.	
   (2001)	
   Folding	
   of	
   intracellular	
   retinol	
   and	
   retinoic	
   acid	
  
binding	
  proteins,	
  Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  43,	
  292-­‐302.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nat.	
  Chem.	
  Biol.	
  10,	
  884-­‐891.	
  

122	
  

70.	
  
	
  
71.	
  
	
  
72.	
  
	
  
73.	
  
	
  
74.	
  
	
  
75.	
  
	
  
76.	
  
	
  
77.	
  
	
  
78.	
  
	
  
79.	
  

	
  
80.	
  

Crist,	
   R.	
   M.,	
   Vasileiou,	
   C.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2006)	
  
Engineering	
  a	
  Rhodopsin	
  Protein	
  Mimic,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  128,	
  
4522-­‐4523.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  J	
  Am	
  Chem	
  Soc	
  135,	
  16111-­‐16119.	
  
Wang,	
  W.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2014)	
  The	
  photochemical	
  determinants	
  of	
  color	
  
vision:	
   revealing	
   how	
   opsins	
   tune	
   their	
   chromophore's	
   absorption	
   wavelength,	
  
BioEssays	
  :	
  news	
  and	
  reviews	
  in	
  molecular,	
  cellular	
  and	
  developmental	
  biology	
  36,	
  65-­‐74.	
  
Wang,	
   W.,	
   Nossoni,	
   Z.,	
   Berbasova,	
   T.,	
   Watson,	
   C.	
   T.,	
   Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Vasileiou,	
   C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   electronic	
   absorption	
   of	
   protein-­‐
embedded	
  all-­‐trans-­‐retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
Nosrati,	
   M.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   A	
  
Photoisomerizing	
  Rhodopsin	
  Mimic	
  Observed	
  at	
  Atomic	
  Resolution,	
  J	
  Am	
  Chem	
  Soc	
  138,	
  
8802-­‐8808.	
  
Berbasova,	
   T.,	
   Santos,	
   E.	
   M.,	
   Nosrati,	
   M.,	
   Vasileiou,	
   C.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2016)	
  Light-­‐Activated	
  Reversible	
  Imine	
  Isomerization:	
  Towards	
  a	
  Photochromic	
  Protein	
  
Switch,	
  Chembiochem	
  :	
  a	
  European	
  journal	
  of	
  chemical	
  biology	
  17,	
  407-­‐414.	
  
Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Berbasova,	
   T.,	
   Nosrati,	
   M.,	
   Jia,	
   X.,	
   Vasileiou,	
   C.,	
   Wang,	
   W.,	
   Santos,	
   E.	
  
M.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2015)	
  "Turn-­‐on"	
  protein	
  fluorescence:	
  in	
  situ	
  formation	
  
of	
  cyanine	
  dyes,	
  J	
  Am	
  Chem	
  Soc	
  137,	
  1073-­‐1080.	
  
Wang,	
  W.,	
  Nossoni,	
  Z.,	
  Berbasova,	
  T.,	
  Watson,	
  C.	
  T.,	
  Yapici,	
  I.,	
  Lee,	
  K.	
  S.	
  S.,	
  Vasileiou,	
  C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   Electronic	
   Absorption	
   of	
   Protein-­‐
Embedded	
  All-­‐trans-­‐Retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
Gunasekaran,	
   K.,	
   Hagler,	
   A.	
   T.,	
   and	
   Gierasch,	
   L.	
   M.	
   (2004)	
   Sequence	
   and	
   structural	
  
analysis	
   of	
   cellular	
   retinoic	
   acid-­‐binding	
   proteins	
   reveals	
   a	
   network	
   of	
   conserved	
  
hydrophobic	
  interactions,	
  Proteins-­‐Structure	
  Function	
  and	
  Genetics	
  54,	
  179-­‐194.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  Crystallogr.	
  Sect.	
  D.	
  Biol.	
  
Crystallogr.	
  70,	
  3226-­‐3232.	
  
Shameer,	
   K.,	
   Shingate,	
   P.	
   N.,	
   Manjunath,	
   S.	
   C.	
   P.,	
   Karthika,	
   M.,	
   Pugalenthi,	
   G.,	
   and	
  
Sowdhamini,	
   R.	
   (2011)	
   3DSwap:	
   curated	
   knowledgebase	
   of	
   proteins	
   involved	
   in	
   3D	
  
domain	
  swapping,	
  Database-­‐the	
  Journal	
  of	
  Biological	
  Databases	
  and	
  Curation.	
  
123	
  

81.	
  

	
  
82.	
  
	
  
83.	
  
	
  
84.	
  
	
  
85.	
  
	
  
86.	
  
	
  
87.	
  
	
  
88.	
  
	
  
89.	
  
	
  
90.	
  
	
  
91.	
  
	
  
92.	
  

Sillitoe,	
   I.,	
   Cuff,	
   A.	
   L.,	
   Dessailly,	
   B.	
   H.,	
   Dawson,	
   N.	
   L.,	
   Furnham,	
   N.,	
   Lee,	
   D.,	
   Lees,	
   J.	
   G.,	
  
Lewis,	
   T.	
   E.,	
   Studer,	
   R.	
   A.,	
   Rentzsch,	
   R.,	
   Yeats,	
   C.,	
   Thornton,	
   J.	
   M.,	
   and	
   Orengo,	
   C.	
   A.	
  
(2013)	
  New	
  functional	
  families	
  (FunFams)	
  in	
  CATH	
  to	
  improve	
  the	
  mapping	
  of	
  conserved	
  
functional	
  sites	
  to	
  3D	
  structures,	
  Nucleic	
  Acids	
  Research	
  41,	
  D490-­‐D498.	
  
Englander,	
   S.	
   W.,	
   and	
   Mayne,	
   L.	
   (2014)	
   The	
   nature	
   of	
   protein	
   folding	
   pathways,	
   Proc	
  
Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  111,	
  15873-­‐15880.	
  
Englander,	
  S.	
  W.,	
  Mayne,	
  L.,	
  and	
  Krishna,	
  M.	
  M.	
  (2007)	
  Protein	
  folding	
  and	
  misfolding:	
  
mechanism	
  and	
  principles,	
  Quarterly	
  reviews	
  of	
  biophysics	
  40,	
  287-­‐326.	
  
Krishna,	
   M.	
   M.,	
   and	
   Englander,	
   S.	
   W.	
   (2007)	
   A	
   unified	
   mechanism	
   for	
   protein	
   folding:	
  
predetermined	
   pathways	
   with	
   optional	
   errors,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  16,	
  449-­‐464.	
  
Mayne,	
   L.,	
   and	
   Englander,	
   S.	
   W.	
   (2000)	
   Two-­‐state	
   vs.	
   multistate	
   protein	
   unfolding	
  
studied	
   by	
   optical	
   melting	
   and	
   hydrogen	
   exchange,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  9,	
  1873-­‐1877.	
  
Roder,	
  H.,	
  Elove,	
  G.	
  A.,	
  and	
  Englander,	
  S.	
  W.	
  (1988)	
  Structural	
  characterization	
  of	
  folding	
  
intermediates	
   in	
   cytochrome	
   c	
   by	
   H-­‐exchange	
   labelling	
   and	
   proton	
   NMR,	
   Nature	
   335,	
  
700-­‐704.	
  
Martins,	
   N.	
   F.,	
   and	
   Santoro,	
   M.	
   M.	
   (1999)	
   Partially	
   folded	
   intermediates	
   during	
  
trypsinogen	
  denaturation,	
  Brazilian	
  journal	
  of	
  medical	
  and	
  biological	
  research	
  =	
  Revista	
  
brasileira	
  de	
  pesquisas	
  medicas	
  e	
  biologicas	
  32,	
  673-­‐682.	
  
Medina,	
   E.,	
   Cordova,	
   C.,	
   Villalobos,	
   P.,	
   Reyes,	
   J.,	
   Komives,	
   E.	
   A.,	
   Ramirez-­‐Sarmiento,	
   C.	
  
A.,	
   and	
   Babul,	
   J.	
   (2016)	
   Three-­‐Dimensional	
   Domain	
   Swapping	
   Changes	
   the	
   Folding	
  
Mechanism	
  of	
  the	
  Forkhead	
  Domain	
  of	
  FoxP1,	
  Biophysical	
  journal	
  110,	
  2349-­‐2360.	
  
Zwanzig,	
   R.	
   (1997)	
   Two-­‐state	
   models	
   of	
   protein	
   folding	
   kinetics,	
   Proceedings	
   of	
   the	
  
National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  94,	
  148-­‐150.	
  
Sanson,	
   B.,	
   Wang,	
   T.,	
   Sun,	
   J.,	
   Wang,	
   L.,	
   Kaczocha,	
   M.,	
   Ojima,	
   I.,	
   Deutsch,	
   D.,	
   and	
   Li,	
   H.	
  
(2014)	
  Crystallographic	
  study	
  of	
  FABP5	
  as	
  an	
  intracellular	
  endocannabinoid	
  transporter,	
  
Acta	
  crystallographica.	
  Section	
  D,	
  Biological	
  crystallography	
  70,	
  290-­‐298.	
  
Assar,	
   Z.,	
   Nossoni,	
   Z.,	
   Wang,	
   W.,	
   Santos,	
   Elizabeth	
  M.,	
   Kramer,	
   K.,	
   McCornack,	
   C.,	
  
Vasileiou,	
  C.,	
  Borhan,	
  B.,	
  and	
  Geiger,	
  James	
  H.	
  Domain-­‐Swapped	
  Dimers	
  of	
  Intracellular	
  
Lipid-­‐Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure.	
  
Silvaroli,	
   J.	
   A.,	
   Arne,	
   J.	
   M.,	
   Chelstowska,	
   S.,	
   Kiser,	
   P.	
   D.,	
   Banerjee,	
   S.,	
   and	
   Golczak,	
   M.	
  
(2016)	
   Ligand	
   Binding	
   Induces	
   Conformational	
   Changes	
   in	
   Human	
   Cellular	
   Retinol-­‐
124	
  

	
  
93.	
  
	
  
94.	
  
	
  
95.	
  

	
  
96.	
  
97.	
  

	
  
98.	
  
	
  
99.	
  

	
  
100.	
  
	
  
101.	
  

	
  

binding	
  Protein	
  1	
  (CRBP1)	
  Revealed	
  by	
  Atomic	
  Resolution	
  Crystal	
  Structures,	
  The	
  Journal	
  
of	
  biological	
  chemistry	
  291,	
  8528-­‐8540.	
  
Gillilan,	
  R.	
  E.,	
  Ayers,	
  S.	
  D.,	
  and	
  Noy,	
  N.	
  (2007)	
  Structural	
  basis	
  for	
  activation	
  of	
  fatty	
  acid-­‐
binding	
  protein	
  4,	
  J	
  Mol	
  Biol	
  372,	
  1246-­‐1260.	
  
Sessler,	
  R.	
  J.,	
  and	
  Noy,	
  N.	
  (2005)	
  A	
  ligand-­‐activated	
  nuclear	
  localization	
  signal	
  in	
  cellular	
  
retinoic	
  acid	
  binding	
  protein-­‐II,	
  Molecular	
  cell	
  18,	
  343-­‐353.	
  
Armstrong,	
  E.	
  H.,	
  Goswami,	
  D.,	
  Griffin,	
  P.	
  R.,	
  Noy,	
  N.,	
  and	
  Ortlund,	
  E.	
  A.	
  (2014)	
  Structural	
  
Basis	
   for	
   Ligand	
   Regulation	
   of	
   the	
   Fatty	
   Acid-­‐binding	
   Protein	
   5,	
   Peroxisome	
   Proliferator-­‐
activated	
   Receptor	
   β/δ	
   (FABP5-­‐PPARβ/δ)	
   Signaling	
   Pathway,	
   The	
   Journal	
   of	
   biological	
  
chemistry	
  289,	
  14941-­‐14954.	
  
Greenfield,	
  N.	
  J.	
  (2006)	
  Using	
  circular	
  dichroism	
  collected	
  as	
  a	
  function	
  of	
  temperature	
  
to	
  determine	
  the	
  thermodynamics	
  of	
  protein	
  unfolding	
  and	
  binding	
  interactions,	
  Nature	
  
Protocols	
  1,	
  2527-­‐2535.	
  
Ramos,	
   C.	
   R.	
   R.,	
   Oyama,	
   S.,	
   Jr.,	
   Sforca,	
   M.	
   L.,	
   Pertinhez,	
   T.	
   A.,	
   Ho,	
   P.	
   L.,	
   and	
   Spisni,	
   A.	
  
(2008)	
   2POA:	
   Schistosoma	
   mansoni	
   Sm14	
   Fatty	
   Acid-­‐Binding	
   Protein:	
   improvement	
   of	
  
protein	
   stability	
   by	
   substitution	
   of	
   the	
   single	
   Cys62	
   residue,	
   Worldwide	
   Protein	
   Data	
  
Bank.	
  
Gill,	
   S.	
   C.,	
   and	
   Vonhippel,	
   P.	
   H.	
   (1989)	
   CALCULATION	
   OF	
   PROTEIN	
   EXTINCTION	
  
COEFFICIENTS	
   FROM	
   AMINO-­‐ACID	
   SEQUENCE	
   DATA,	
   Analytical	
   Biochemistry	
   182,	
   319-­‐
326.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Davis,	
  I.	
  W.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  
J.,	
   Hung,	
   L.	
   W.,	
   Kapral,	
   G.	
   J.,	
   Grosse-­‐Kunstleve,	
   R.	
   W.,	
   McCoy,	
   A.	
   J.,	
   Moriarty,	
   N.	
   W.,	
  
Oeffner,	
  R.,	
  Read,	
  R.	
  J.,	
  Richardson,	
  D.	
  C.,	
  Richardson,	
  J.	
  S.,	
  Terwilliger,	
  T.	
  C.,	
  and	
  Zwart,	
  
P.	
   H.	
   (2010)	
   PHENIX:	
   a	
   comprehensive	
   Python-­‐based	
   system	
   for	
   macromolecular	
  
structure	
  solution,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  213-­‐
221.	
  
Bailey,	
  S.	
  (1994)	
  THE	
  CCP4	
  SUITE	
  -­‐	
  PROGRAMS	
  FOR	
  PROTEIN	
  CRYSTALLOGRAPHY,	
  Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  50,	
  760-­‐763.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  J.,	
  Hung,	
  L.-­‐
W.,	
  Jain,	
  S.,	
  Kapral,	
  G.	
  J.,	
  Kunstleve,	
  R.	
  W.	
  G.,	
  McCoy,	
  A.	
  J.,	
  Moriarty,	
  N.	
  W.,	
  Oeffner,	
  R.	
  
D.,	
   Read,	
   R.	
   J.,	
   Richardson,	
   D.	
   C.,	
   Richardson,	
   J.	
   S.,	
   Terwilliger,	
   T.	
   C.,	
   and	
   Zwart,	
   P.	
   H.	
  
(2011)	
  The	
  Phenix	
  software	
  for	
  automated	
  determination	
  of	
  macromolecular	
  structures,	
  
Methods	
  55,	
  94-­‐106.	
  

125	
  

102.	
  
	
  
103.	
  
	
  
104.	
  
	
  
105.	
  
	
  
106.	
  
	
  
	
  
	
  

Murshudov,	
  G.	
  N.,	
  Vagin,	
  A.	
  A.,	
  and	
  Dodson,	
  E.	
  J.	
  (1997)	
  Refinement	
  of	
  macromolecular	
  
structures	
   by	
   the	
   maximum-­‐likelihood	
   method,	
   Acta	
   Crystallographica	
   Section	
   D-­‐
Biological	
  Crystallography	
  53,	
  240-­‐255.	
  
Ong,	
   D.	
   E.,	
   and	
   Page,	
   D.	
   L.	
   (1987)	
   Cellular	
   retinol-­‐binding	
   protein	
   (type	
   two)	
   is	
   abundant	
  
in	
  human	
  small	
  intestine,	
  Journal	
  of	
  Lipid	
  Research	
  28,	
  739-­‐745.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins	
   –	
   insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  Lipid	
  Research	
  43,	
  328-­‐349.	
  
Gillilan,	
   R.	
   E.,	
   Ayers,	
   S.	
   D.,	
   and	
   Noy,	
   N.	
   (2007)	
   Structural	
   Basis	
   for	
   Activation	
   of	
   Fatty	
  
Acid-­‐binding	
  Protein	
  4,	
  Journal	
  of	
  Molecular	
  Biology	
  372,	
  1246-­‐1260.	
  
Kaczocha,	
   M.,	
   Glaser,	
   S.	
   T.,	
   and	
   Deutsch,	
   D.	
   G.	
   (2009)	
   Identification	
   of	
   intracellular	
  
carriers	
  for	
  the	
  endocannabinoid	
  anandamide,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  106,	
  6375-­‐6380.	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  

126	
  

Chapter	
  IV:	
  STRUCTURAL	
  INSIGHTS	
  INTO	
  EMISSION	
  REGULATION	
  OF	
  HCRBBPII-­‐BOUND	
  
SOLVATOCHROMIC	
  FLUOROPHORE	
  
	
  

IV-­‐1	
  Introduction	
  
As	
  one	
  part	
  of	
  the	
  collaboration	
  (described	
  in	
  chapter	
  II)	
  between	
  Prof.	
  Borhan	
  and	
  Prof.	
  
Geiger	
   lab,	
   hCRBPII	
   was	
   reengineered	
   so	
   that	
   it	
   can	
   bind	
   to	
   the	
   fluorophores	
   to	
   make	
   a	
   new	
  
system	
  with	
  the	
  overreaching	
  goal	
  of	
  using	
  these	
  systems	
  as	
  fluorescent	
  protein	
  tags	
  over	
  the	
  
entire	
  visible	
  spectrum.	
  (75)	
  
This	
  new	
  class	
  of	
  fluorogenic	
  proteins	
  was	
  created	
  by	
  binding	
  fluorophore	
  aldehydes	
  in	
  
the	
  binding	
  pocket	
  of	
  hCRBPII	
  via	
  protonated	
  Schiff	
  base	
  (PSB)	
  formation,	
  which	
  as	
  a	
  result	
  is	
  
sensitive	
  to	
  pH.	
   (111)	
  Numerous	
  applications	
  can	
  be	
  intended	
  such	
  as	
  application	
  in	
  multi-­‐color	
  
imaging	
  and	
  in	
  vivo	
  pH	
  sensing.(112)	
  One	
  of	
  the	
  other	
  benefits	
  of	
  this	
  system	
  is	
  that	
  no	
  oxygen	
  is	
  
required	
  (unlike	
  green	
  fluorescent	
  protein,	
  GFP)	
  and	
  therefore	
  it	
  could	
  find	
  potential	
  in	
  obligate	
  
anaerobes.	
   Dr.	
   Santos	
   in	
   prof.	
   Borhan’s	
   group	
   has	
   synthesized	
   a	
   new	
   solvatochromic	
  
fluorophore,	
  ThioFlour,	
  and	
  developed	
  this	
  fluorogenic	
  system.	
  	
  (Figure	
  IV-­‐1c)	
  
Solvatochromism	
   describes	
   the	
   alteration	
   in	
   wavelength	
   (and	
   intensity),	
   of	
   an	
  
absorption	
  and/or	
  emission	
  band	
  that	
  comes	
  from	
  the	
  polarity	
  change	
  of	
  its	
  medium.(113),	
   (114)	
  
Since	
   specificity	
   is	
   crucial	
   in	
   engineering	
   a	
   practical	
   and	
   efficient	
   fluorogenic	
   system,	
  
solvatochromic	
   fluorophores	
   were	
   chosen	
   to	
   couple	
   with	
   the	
   hCRBPII	
   variants.	
   Also	
   previous	
  
studies	
  with	
  a	
  merocyanine	
  aldehyde	
  showed	
  non-­‐specific	
  background	
  fluorescence	
  in	
  imaging	
  
of	
   live	
   cells	
   (arising	
   from	
   the	
   iminium	
   formation	
   with	
   nonspecific	
   proteins),	
   due	
   to	
   the	
   static	
  
excitation	
  and	
  emission	
  profiles	
  of	
  these	
  fluorophore	
  bound	
  hCRBPII	
  complexes.	
  (111)	
  

127	
  

In	
   this	
   new	
   system,	
   emission	
   of	
   solvatochromic	
   fluorophores	
   is	
   flexible	
   based	
   on	
   the	
  
polarity	
   of	
   their	
   environment.	
   Therefore	
   multicolor	
   probes	
   can	
   be	
   developed.	
   More	
  
importantly,	
   due	
   to	
   possibility	
   of	
   the	
   hCRBPII/fluorophore	
   emission	
   tuning,	
   we	
   can	
   reduce	
  
nonspecific	
  labeling.	
  This	
  would	
  result	
  in	
  a	
  probe	
  that	
  yields	
  little	
  background	
  fluorescence.	
  

IV-­‐2	
  Rational	
  behind	
  the	
  parent	
  template:	
  Q108K:K40L:T51V:T53S	
  
Most	
   solvatochromic	
   dyes	
   show	
   relatively	
   low	
   fluorescence	
   quantum	
   yields	
   in	
   polar	
  
solvents	
   and	
   high	
   fluorescence	
   quantum	
   yields	
   in	
   apolar	
   solvents.(114),(115)	
   	
   ThioFlour	
  
spectroscopic	
  data	
  in	
  various	
  solvents	
  suggests	
  the	
  same	
  result.	
  (Performed	
  by	
  Dr.	
  Santos,	
  Data	
  
not	
  shown)	
  	
  It	
  also	
  shows	
  no	
  fluorescent	
  in	
  PBS	
  solutions	
  (can	
  be	
  considered	
  as	
  no	
  fluorescence	
  
being	
   observed	
   of	
   unbound	
   probe	
   during	
   in	
   vivo	
   applications).	
   This	
   could	
   be	
   beneficial	
   in	
   live	
  
cell	
   imaging.	
   	
   Her	
   studies	
   also	
   show	
   that	
   ThioFluor-­‐PSB	
   is	
   much	
   less	
   solvatochromic	
   than	
   the	
  
ThioFluor	
   aldehyde,	
   indicating	
   that	
   PSB	
   emission	
   does	
   not	
   follow	
   the	
   same	
   solvatochromic	
  
trend	
  and	
  some	
  other	
  factors,	
  such	
  as	
  hydrogen	
  bonding,	
  may	
  affect	
  the	
  emission	
  wavelength	
  
of	
  ThioFlour-­‐PSB.(116)	
  
To	
   peruse	
   the	
   study	
   of	
   ThioFluor	
   bound	
   protein	
   interactions,	
   Dr.	
   Santos	
   generated	
   a	
  
large	
  number	
  of	
  hCRBPII	
  variants	
  by	
  point	
  mutation.	
  Table	
   IV-­‐1	
  shows	
  only	
  the	
  ones	
  to	
  being	
  
discussed	
   in	
   this	
   chapter.	
   For	
   this	
   purpose	
   the	
   polarity	
   can	
   be	
   removed	
   in	
   two	
   ways:	
   	
   1)	
  
Removal	
   of	
   the	
   more	
   polar	
   residues	
   in	
   the	
   binding	
   pocket	
   nearest	
   the	
   chromophore	
   and	
   2)	
  
Removal	
  of	
  the	
  direct	
  water	
  molecules	
  interacting	
  with	
  the	
  iminium	
  region.	
  Removal	
  of	
  these	
  
negative	
  dipoles	
  in	
  the	
  PSB	
  region	
  should	
  encourage	
  positive	
  charge	
  delocalization.	
  (76)	
  	
  
As	
  previously	
  mentioned	
  in	
  Chapter	
  II,	
  hCRBPII	
  has	
  a	
  large	
  hydrophobic	
  binding	
  pocket,	
  
with	
   few	
   polar	
   residues.	
   (76)	
   Mutagenesis	
   was	
   first	
   performed	
   on	
   T51,	
   T53,	
   Q4,	
   Q38	
   and	
   R58	
  
128	
  

residues	
   (Figure	
   IV-­‐1),	
   mutating	
   them	
   to	
   similarly	
   sized	
   hydrophobic	
   residues	
   in	
   the	
   parent	
  
double	
   mutant	
   Q108K:K40L	
   (KL)	
   form.	
   KL	
   was	
   used.	
   Q108K	
   couldn’t	
   be	
   used	
   due	
   to	
   its	
  
precipitation	
   in	
   acidic	
   pH.	
  

(76)

	
   Additionally,	
   we	
   only	
   discuss	
   the	
   monomer	
   spectroscopic	
  

properties	
  here.	
  Since	
  T51V	
  is	
  capable	
  of	
  monomer	
  formation	
  of	
  hCRBPII	
  variants,	
  it	
  was	
  thus	
  
retained	
  in	
  most	
  mutants	
  tested.	
  	
  
	
  
a.	
  

	
  

R58

Q38

T53

T51

	
  

Q4

	
  

Q108K

	
  
	
  

b.	
  

c.	
  
Re3nal

	
  
	
  

ThioFLuor

ThioFluor

Q108K

	
  
Figure	
   IV-­‐1:	
   a.	
   Polar	
  residues	
  around	
  the	
  ligand	
   mutated	
  in	
  order	
  to	
   remove	
  polarity	
  from	
  
the	
  binding	
  pocket.	
  Coordinates	
  obtained	
  from	
  PDB	
  4EXZ	
  (hCRBPII-­‐Q108K:K40L/retinal).	
  	
  b.	
  
Both	
   retinal	
   and	
   ThioFluor	
   bind	
   and	
   lay	
   in	
   the	
   same	
   region	
   of	
   the	
   binding	
   ppocket.	
   (the	
  
Thioflour	
   coordinates	
   are	
   from	
   	
   Q108K:K40L:	
   T51V:T53S:	
   R58Y	
   structure,	
   will	
   be	
   discussed	
  
later	
  in	
   the	
  IV-­‐3	
  section.	
   c.	
   The	
   chemical	
  Structure	
   of	
   the	
  ThioFluor	
   ,	
   which	
   	
   was	
  synthesized	
  
by	
  Dr.	
  Santos	
  
	
  
T53	
   is	
   located	
   near	
   the	
   middle	
   of	
   the	
   polyene	
   in	
   the	
   crystal	
   structure	
   of	
   KL/retinal,	
   as	
  
shown	
  in	
  Figure	
   IV-­‐2a	
   making	
  a	
  hydrogen	
  bond	
  to	
  T51	
  through	
  a	
  water	
  molecule.	
  In	
  order	
  to	
  
alter	
   its	
   polarity,	
   T53	
   was	
   replaced	
   with	
   various	
   residues	
   including	
   alanine	
   and	
   serine	
   in	
   the	
  
parent	
  template	
  KL.	
  
	
  
129	
  

	
  Interestingly,	
   all	
   these	
   substitutions	
   led	
   to	
   at	
   least	
   a	
   slight	
   red	
   shift	
   in	
   absorption.	
   From	
  
KL	
  the	
  change	
  in	
  absorption	
  was	
  5	
  nm,	
  7	
  nm	
  (3nm	
  and	
  8nm	
  in	
  their	
  emission	
  wavelength)	
  for	
  
T53A	
  and	
  T53S	
  respectively	
  (Table	
  IV-­‐1).	
  	
  As	
  it	
  was	
  mentioned	
  before,	
  T51	
  and	
  T53	
  interact	
  via	
  a	
  
water-­‐mediated	
  hydrogen	
  bond	
  in	
  the	
  crystal	
  structure	
  (of	
  KL	
  bound	
  retinal,	
  Figure	
   IV-­‐2a).	
  By	
  
mutating	
  T51	
  to	
  valine,	
  this	
  interaction	
  would	
  be	
  removed,	
  leading	
  to	
  a	
  red	
  shift	
  in	
  wavelength.	
  
T53S	
  shows	
  a	
  similar	
  effect.	
  There	
  is	
  a	
  water	
  network	
  between	
  Q38	
  and	
  Q128	
  (Figure	
   IV-­‐2b).	
  
T53S	
   could	
   potentially	
   form	
   a	
   hydrogen	
   bond	
   with	
   Q38,	
   which	
   would	
   disrupt	
   the	
   conserved	
  
water-­‐mediated	
  interaction	
  between	
  Q38	
  and	
  Q128	
  by	
  ejection	
  of	
  water,	
  and	
  thus	
  change	
  the	
  
polarity,	
  near	
  the	
  ThioFluor.	
  	
  
	
  
	
  

a.	
  

b.	
  

	
  

Q128

3.0A%
3.1A%

	
  

2.9A%
2.8A%

	
  
Q38

Q108K

	
  

Figure	
   IV-­‐2:	
   a.	
   Water	
   network	
  hydrogen	
  bonding	
  between	
  T51	
   and	
  T53.	
   b.	
   Water	
   mediated	
  
hydrogen	
   bonding	
   between	
   Q38	
   and	
   Q128.	
   (PDB	
   4EXZ,	
   hCRBPII-­‐Q108K:K40L/retinal,	
   Retinal	
   is	
  
represented	
  in	
  green	
  sticks).	
  	
  
	
  
To	
   test	
   this	
   hypothesis,	
   Dr.	
   Santos	
   generated	
   Q108K:K40L:T53A:R58F.	
   This	
   mimics	
   the	
   T53S	
  
mutation	
   and	
   elimination	
   of	
   the	
   water	
   mediated	
   interaction	
   between	
   Q38	
   and	
   Q128	
   leading	
   to	
  
removing	
   polarity	
   in	
   the	
   pocket.	
   	
   In	
   case	
   of	
   residue	
   58	
   (R58),	
   it	
   was	
   mutated	
   to	
   aromatic	
  
residues,	
  tyrosine	
  and	
  phenylalanine	
  in	
  KL	
  and	
  both	
  led	
  to	
  a	
  bathochromic	
  shift	
  in	
  absorption	
  
130	
  

11nm	
  and	
  17nm	
  and	
  emission	
  7	
  and	
  4nm	
  with	
  ThioFluor	
  respectively	
  (Table	
   IV-­‐1).	
  This	
  result	
  
agrees	
   with	
   previous	
   studies	
   that	
   showed	
   that	
   isolating	
   the	
   retinal	
   chromophore	
   in	
   the	
   binding	
  
pocket	
  of	
  hCRBPII	
  resulted	
  in	
  a	
  bathochromic	
  shift	
  due	
  to	
  more	
  effective	
  shielding	
  	
  
of	
  the	
  binding	
  pocket.

	
  

(75, 76, 112)

It	
  should	
  be	
  noted	
  that	
  residues	
  closer	
  to	
  the	
  iminium,	
  Q4	
  and	
  T51,	
  led	
  to	
  a	
  larger	
  red-­‐shift	
  than	
  
T53	
  and	
  Q38	
  residues	
  farther	
  away.	
  Therefore,	
  the	
  template	
  Q108K:K40L:T51V:T53S	
  was	
  used	
  
for	
  further	
  protein	
  engineering	
  by	
  Dr.	
  Santos.	
  However	
  Q4	
  mutations	
  were	
  not	
  included,	
  due	
  to	
  
prior	
   studies	
   showing	
   that	
   the	
   pKa	
   of	
   hCRBPII/retinal	
   complexes	
   are	
   severely	
   low	
   when	
   Q4	
   was	
  
mutated.	
  (76)	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

131	
  

!

Table	
  IV-­‐1:	
  Spectroscopic	
  change	
  of	
  hCRBPII	
  variants	
  bound	
  ThioFluor	
  
Entry& Mutant&
1&
Q108K:K40L&
&
2&
Q108K:K40L:T51V&
&
3&
Q108K:K40L:T53A&
&
4&
Q108K:K40L:T53S&
&
5&
Q108K:K40L:&T51V:T53S&
&
6&
Q108K:K40L:Q38F&
&
7&
Q108K:K40L:&T51V:T53S:&R58H&
&
8&
Q108K:K40L:&T51V:T53S:&R58W&
&
9&
Q108K:K40L:&T51V:T53S:&R58Y&
&
10&
Q108K:K40L:&T53A:&R58F&
&
11&
Q108K:K40L:&T51V:T53S:&R58W:Y19W&
&
12&
Q108K:K40L:&T51V:T53S:&R58W:&Y19W:&
A33W&
&
13&
Q108K:K40L:&T51V:T53S:&
R58W:Y19W:L117E&
&
14&
Q108K:K40L:&T51V:T53S:&R58W:Y19W:&
A33W:&L117E&
&

Îťabs&
580&

Îťem &
674&

ϕ&
nd&

608&

690&

0.12&

585&

677&

0.11&

587&

682&

0.08&

627&

697&

0.15&

588&

680&

0.13&

643&&
&
623&

705&

0.14&

697&

0.08&

622&

691&

0.11&

589&

681&

0.12& 	
  

653&

719&

0.10& 	
  

666&

724&

0.14& 	
  

558&
539&

673&
628&

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

0.31&
0.28&

	
  
	
  

IV-­‐3	
  Attempts	
  to	
  isolate	
  the	
  binding	
  cavity	
  	
  
Due	
  to	
  previous	
   studies	
  conducted	
  by	
  Dr.	
  Wenjing	
  Wang	
  and	
  Dr.	
  Nossoni,	
  an	
  aromatic	
  
residue	
   at	
   position	
   R58	
   leads	
   to	
   a	
   bathochromic	
   shift	
   in	
   absorption	
   when	
   retinal	
   is	
   used	
   as	
   a	
  
ligand,	
   due	
   to	
   more	
   effective	
   shielding	
   of	
   the	
   binding	
   poc	
   ket.	
   (75,	
   76,	
   112)	
   (117)Therefore	
   R58F,	
  
R58Y,	
   R58W	
   and	
   R58H	
   mutations	
   were	
   introduced	
   into	
   the	
   parent	
   mutant	
  

132	
  

Q108K:K40L:T51V:T53S	
   by	
   Dr.	
   Santos	
   in	
   this	
   project.	
   It	
   should	
   be	
   noted	
   that,	
   unlike	
   retinal,	
  
ThioFluor	
  contains	
  heteroatoms	
  that	
  can	
  interact	
  with	
  surrounding	
  residues	
  through	
  hydrogen	
  
bonding,	
  hydrophobic	
  interactions,	
  pi-­‐pi	
  stacking,	
  pi-­‐cation	
  interactions,	
  ionic	
  interactions,	
  etc.	
  
Crystal	
   Structure	
   of	
   Q108K:K40L:	
   T51V:T53S:	
   R58W/ThioFLuor	
   was	
   obtioned	
   by	
   Dr.	
   Nosrati.	
  
Based	
   on	
   its	
   structure,	
   R58W	
   sits	
   directly	
   on	
   top	
   of	
   the	
   N,N-­‐dimethyl	
   amino	
   moiety	
   of	
   the	
  
bound	
  ThioFluor.	
  As	
  shown	
  (Figure	
  IV-­‐3),	
  the	
  distance	
  between	
  the	
  tryptophan	
  and	
  nitrogen	
  is	
  
3.7	
  Å,	
  well	
  within	
  van	
  der	
  Waals’	
  contact.	
  	
  
	
  

a.	
  

Q38

b.	
  

R58W

	
  

T53S

R58W

T51V

	
  

2.7A/

2.9A/

3.7A.

2.7A/

T53S

	
  

Q128

Q108K

Q108K

	
  

T51V

	
  
	
  
	
  
	
   	
  

Figure	
   IV-­‐3:	
   a.	
   Crystal	
   structure	
   of	
   Q108K:K40L:T51V:T53S:R58W/ThioFluor	
   and	
  
space-­‐filling	
  representation	
  obtained	
  by	
  Dr.	
  Nosrati.	
  R58W	
  Is	
  packed	
  with	
  the	
  N,N-­‐
dimethyl	
   amino	
   moiety	
   of	
   the	
   ThioFlour	
   tail.	
   b.	
   The	
   water	
   mediated	
   hydrogen	
  
bonding	
  between	
  T53S	
  and	
  R58W	
  is	
  highlighted.	
  	
  
	
  

In	
   that	
   series	
   R58H	
   mutation	
   caused	
   16nm	
   in	
   absorption	
   and	
   8nm	
   in	
   emission	
   from	
  

Q108K:K40L:T51V:T53S.	
  Unfortunately,	
  crystallization	
  of	
  Q108K:K40L:T51V:T53S:R58H	
  was	
  not	
  
successful.	
  
In	
   the	
   R58Y	
   mutation	
   series,	
   tyrosine	
   is	
   flipped	
   out	
   of	
   the	
   pocket,	
   providing	
   no	
   stacking	
  
interaction	
   with	
   the	
   ligand	
   (which	
   yielded	
   a	
   much	
   smaller	
   wavelength	
   alteration	
   than	
  
Q108K:K40L:T51V:T53S.)	
   Alternatively,	
   the	
   R58Y	
   participates	
   in	
   a	
   water-­‐mediated	
   hydrogen	
  
133	
  

bond	
  with	
  the	
  hydroxyl	
  side	
  chain	
  of	
  T29	
  (Figure	
   IV-­‐4).	
  As	
  the	
  ligand	
  is	
  not	
  locked	
  in	
  place	
  by	
  
the	
  R58	
  mutation,	
  this	
  presumably	
  causes	
  the	
  translocation	
  and	
  rotation	
  of	
  the	
  chromophore.	
  	
  
	
  
a.	
  

b.	
  

	
  

R58Y
R58Y

R58Y

	
  
	
  

2.7A%

	
  

3.1A%

	
  

T29

	
  
	
  
	
  
	
  
	
  
	
  

Figure	
  IV-­‐4:	
  a.	
  Overlay	
  of	
  Crystal	
  structures	
  of	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  
(cyan)	
  with	
  Q108K:K40L:T51V:T53S:R58Y/ThioFluor	
  (green).	
  b.	
  It	
  shows	
  Water	
  mediated	
  
hydrogen	
  bonding	
  between	
  R58Y	
  and	
  T29.	
  
	
  

Between	
   Q108K:K40L:T51V:T53S:R58W/ThioFluor	
   and	
   Q108K:K40L:T51V:T53S:R58Y/ThioFluor,	
  
ligand	
   trajectories	
   are	
   perpendicular	
   ,	
   both	
   have	
   cis	
   iminium	
   bonds.	
   However	
   R58Y	
   makes	
   a	
  
hydrogen	
   bonding	
   network	
   through	
   a	
   water	
   molecule	
   and	
   Thr29.	
   On	
   the	
   other	
   hand,	
   Trp	
   58	
  
packed	
   with	
   the	
   tail	
   of	
   ligand	
   and,	
   makes	
   a	
   pi	
   cation	
   interaction	
   with	
   the	
   ligand,	
   therefore	
  
causing	
  red	
  shift	
  emission	
  for	
  R58W.	
  
Most	
  

remarkably,	
  

upon	
  

overlay	
  

of	
  

the	
  

structure	
  

with	
  

both	
  

Q108K:K40L:T51V:T53S:R58W/ThioFluor	
   and	
   Q108K:K40L:T51V:T53S:R58Y/ThioFluor,	
   the	
  
flexibility	
   of	
   the	
   loop	
   at	
   the	
   entrance	
   of	
   the	
   binding	
   pocket	
   can	
   be	
   seen	
   (Figure	
  IV-­‐4).	
   Although	
  

134	
  

	
  

the	
  crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58F/ThioFluor	
  was	
  not	
  obtained,	
  the	
  structure	
  
of	
  the	
  similar	
  variant	
  Q108K:K40L:T53A:R58F	
  with	
  ThioFluor	
  was	
  solved.	
  (Figure	
  IV-­‐5)	
  	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

Figure	
  IV-­‐5:	
  Overlay	
  of	
  Crystal	
  structures	
  of	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  (cyan)	
  
with	
  Q108K:K40L:T51V:T53S:R58Y/ThioFluor	
  (green)	
  and	
  Q108K:K40L:T53A:R58F/ThioFluor	
  
(salmon).	
  	
  
As	
   observed,	
   with	
   an	
   R58F	
   mutation,	
   the	
   flexible	
   loop	
   allows	
   movement	
   such	
   that	
  

phenylalanine	
  is	
  within	
  van	
  der	
  Waals	
  contact	
  of	
  the	
  ligand.	
  A	
  slight	
  bathochromic	
  shift	
  of	
  4	
  nm	
  
in	
  absorption	
  is	
  observed	
  for	
  Q108K:K40L:T53A:R58F	
  from	
  Q108K:K40L:T53A,	
  suggesting	
  again	
  
that	
   R58F	
   does	
   not	
   successfully	
   sequester	
   the	
   binding	
   cavity.	
   (Table	
   IV-­‐1	
   and	
   Figure	
   IV-­‐5)	
   By	
  
comparing	
   Q108K:K40L:T51V:T53S:R58Y/ThioFluor	
   and,	
   Q108K:K40L:T53A:R58F/ThioFluor	
  
structures,	
  	
  a	
  180°	
  difference	
  in	
  ligand	
  trajectory	
  has	
  been	
  revealed.	
  In	
  R58F	
  due	
  to	
  the	
  position	
  
of	
   Phe	
   58,	
   the	
   benzene	
   ring	
   of	
   the	
   ligand	
   has	
   moved	
   lower	
   compared	
   to	
   R58Y.(Figure	
   IV-­‐5)	
  
Further	
   crystallographic	
   analysis	
   would	
   be	
   required	
   in	
   order	
   to	
   determine	
   the	
   effect	
   of	
   other	
  

135	
  

Arg58	
  mutations	
  as	
  they	
  are	
  shown	
  to	
  dramatically	
  alter	
  the	
  chromophore	
  orientation.	
  	
  

IV-­‐4	
   Extensive	
   packing	
   of	
   the	
   ligand	
   and	
   removal	
   of	
   water	
   leads	
   to	
   bathochromic	
   shift	
  
in	
  absorption	
  and	
  emission.	
  
	
  
Based	
  on	
  the	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W/ThioFluor,	
  there	
  is	
  a	
  conserved	
  water	
  

network	
   from	
   the	
   phenol	
   of	
   Tyr	
   19	
   to	
   the	
   thiophene	
   sulfur	
   of	
   ThioFluor	
   (Figure	
   IV-­‐6)	
   As	
  
previously	
  shown,	
  removal	
  of	
  polarity	
  yields	
  to	
  red-­‐shift	
  wavelength.	
  Y19W	
  mutation	
  (less	
  likely	
  
to	
  provide	
  hydrogen	
  bonding,	
  conducted	
  by	
  Dr.	
  Santos)	
  would	
  result	
  in	
  removal	
  of	
  this	
  water	
  
network.	
   	
   A	
   red	
   shift	
   in	
   absorption	
   of	
   30	
   nm	
   is	
   observed	
   with	
   the	
   introduction	
   of	
   Y19W	
   to	
  
Q108K:K40L:T51V:T53S:R58W	
  (Table	
  IV-­‐1).	
  	
  	
  
	
  
	
  
	
  
	
  
	
  

Figure	
   IV-­‐6:	
   The	
   water	
   mediated	
   hydrogen	
   bonding	
   between	
   Tyr19	
   to	
   the	
  
thiophene	
   sulfur	
   of	
   ThioFluor	
   highlighted	
   in	
   the	
   Crystal	
   structure	
   of	
  
Q108K:K40L:T51V:T53S:R58W/ThioFluor	
  obtained	
  by	
  Dr.	
  Nosrati.	
  	
  
	
  
	
  
To	
  

confirm	
  

this	
  

hypothesis,	
  

the	
  

crystal	
  

structure	
  

of	
  

Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor	
   was	
   determined,	
   and	
   as	
   expected	
   shows	
   that	
  
the	
   water	
   network	
   was	
   indeed	
   abolished,	
   abrogating	
   the	
   water-­‐mediated	
   interation	
   with	
   the	
  
fluorophore	
   (Figure	
   IV-­‐7).	
   Moreover,	
   the	
   hydroxyl	
   of	
   Ser	
   53	
   hydrogen	
   bonds	
   with	
   both	
   the	
  

136	
  

indole	
  side	
  chain	
  of	
  Trp	
  58	
  and	
  the	
  amide	
  moiety	
  of	
  Gln	
  38	
  through	
  a	
  water	
  mediated	
  hydrogen	
  
bond	
   (Figure	
   IV-­‐3b).	
   These	
   interactions	
   presumably	
   dictate	
   the	
   chromophore’s	
   orientation.	
  
Between	
   Q108K:K40L:T51V:T53S:R58W	
   and	
   Q108K:K40L:T51V:T53S:R58W:Y19W,	
   the	
   only	
  
difference	
   is	
   the	
   Y19W	
   mutation,	
   which	
   removes	
   the	
   water	
   network	
   and	
   makes	
   a	
   pi	
   stacking	
  
interaction	
   with	
   the	
   ligand,	
   encapsulating	
   the	
   fluorophore	
   and	
   inducing	
   a	
   vast	
   red	
   shift	
   in	
  
emission.	
  (Figure	
  IV-­‐7	
  and	
  9)	
  

a.	
  

	
  

b.	
  

	
  
	
  
	
  
	
  
	
  

Figure	
  IV-­‐7:	
  a.	
  Crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor.	
  b.	
  
Water	
  network	
  is	
  abolished	
  in	
  this	
  structure.	
  	
  ThioFluor	
  and	
  residues	
  around	
  it	
  are	
  
represented	
  in	
  sticks.	
  
	
  
The	
  last	
  attempt	
  in	
  this	
  series	
  was	
  to	
  encapsulate	
  the	
  binding	
  cavity,	
  via	
  the	
  introduction	
  
of	
  one	
  more	
  bulky	
  tryptophan	
  residue	
  at	
  A33,	
  Q108K:K40L:T51V:T53S:R58W:Y19W	
  (Figure	
   IV-­‐
7b).	
   Unfortunately,	
   crystallization	
   of	
   Q108K:K40L:T51V:T53S:R58W:Y19W:A33W	
   was	
   not	
  
successful.	
   Presumably,	
   A33W	
   adequately	
   covers	
   the	
   binding	
   cavity.	
   (See	
   Figure	
   IV-­‐8	
   for	
   a	
  
similar	
  

mutant)	
  

While	
  

efforts	
  

to	
  

crystallize	
  

Q108K:K40L:T51V:T53S:R58W:Y19W:A33W/ThioFluor	
  were	
  not	
  successful,	
  a	
  crystal	
  structure	
  of	
  
137	
  

Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor	
  

was	
  

achieved	
  

by	
  

Alireza	
  

Ghanbarpour.	
   (Figure	
   IV-­‐8	
   and	
   9)	
   It	
   shows	
   that	
   A33W	
   closes	
   the	
   lid	
   of	
   the	
   binding	
   pocket.	
  
Combination	
  of	
  R58W,	
  Y19W	
  and	
  A33W	
  encapsulates	
  the	
  binding	
  cavity.	
  
	
  
	
  
	
  
	
  
	
  

	
  

	
  
	
  
Figure	
  IV-­‐8:	
  Crystal	
  structure	
  of	
  Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor	
  
obtained	
  by	
  Alireza.	
  	
  ThioFluor	
  in	
  cyan	
  and	
  residues	
  around	
  it	
  in	
  green	
  are	
  	
  shown	
  in	
  sticks.	
  

IV-­‐5	
  Localization	
  of	
  charge	
  leads	
  to	
  a	
  blue	
  shift	
  in	
  absorption	
  and	
  emission	
  	
  
Previouosly	
  with	
  hCRBPII/retinal,	
  Dr.	
  Wang	
  and	
  Dr.	
  Nossoni	
  demonstrated	
  that	
  an	
  even	
  
distribution	
   of	
   electrostatic	
   potential	
   across	
   the	
   polyene	
   led	
   to	
   a	
   bathochromic	
   shift	
   in	
  
absorption,	
  and	
  by	
  localizing	
  the	
  cation	
  on	
  the	
  iminium	
  nitrogen	
  it	
  led	
  to	
  a	
  blue	
  shift.	
  Similarly,	
  
we	
  sought	
  to	
  introduce	
  negatively	
  polarized	
  residues	
  near	
  the	
  iminium.	
  (Figure	
  IV-­‐9)	
  

138	
  

	
  

K40L

	
  

Q38
T51V

T53S

R58W

	
  
	
  

A33

	
  
	
  

ThioFluor

Q108K

L117

Y19W

	
  

Figure	
   IV-­‐9:	
   Mutated	
   residues	
   which	
   discussed	
   in	
   this	
   chapter	
   (show	
   in	
   blue)	
   in	
   an	
   attempt	
   to	
  
encapsulate	
  
the	
  
binding	
  
cavity.	
  
Crystal	
  
structure	
  
is	
  
of	
  
Q108K:K40L:T51V:T53S:R58W:Y19W/ThioFluor.	
  	
  
To	
  

this	
  

end,	
  

Leu117	
  

was	
  

mutated	
  

to	
  

glutamic	
  

acid	
  

in	
  

Q108K:K40L:T51V:T53S:R58W:Y19W	
   and	
   Q108K:K40L:T51V:T53S:R58W:Y19W:A33W,	
   both	
   of	
  
which	
   when	
   bound	
   with	
   ThioFluor	
   producing	
   a	
   clearly	
   observable	
   red-­‐shifted	
   iminium.	
  
(Performed	
  by	
  Dr.	
  Santos)	
  
Between	
  

Structures	
  

of	
  

Q108K:K40L:T51V:T53S:R58W:Y19W	
  

and	
  

Q108K:K40L:T51V:T53S:R58W:Y19W:L117E	
  they	
  have	
  cis	
  and	
  trans	
  iminium	
  bond	
  respectively.	
  
L11E	
   mutant	
   makes	
   hydrogen	
   bonds	
   with	
   iminnium	
   bond	
   and	
   localizes	
   the	
   positive	
   charge	
   of	
  
iminium.	
  It	
  also	
  gives	
  the	
  imine	
  trans	
  over	
  the	
  cis	
  isomer.	
  	
  Additionally,	
  the	
  entire	
  ligand	
  moved	
  

139	
  

in	
   the	
   horizontal	
   direction	
   as	
   well.	
   This	
   explains	
   why	
   the	
   carboxylate	
   blue	
   shifts	
   100nm	
   and	
  
46nm	
  for	
  absorption	
  and	
  emission	
  respectively.	
  (Figure	
  IV-­‐14)	
  	
  

	
  
Figure	
  
IV-­‐10:	
  Crystal	
  structure	
  of	
  
Q108K:K40L:T51V:T53S:R58W:Y19W:A33W:L117E/ThioFluor.	
  	
  ThioFluor	
  and	
  
highlighted	
  residues	
  are	
  represented	
  as	
  sticks.	
  
	
  
As	
  other	
  variants	
  in	
  these	
  series,	
  we	
  can	
  point	
  out	
  Q108K:K40L:T53A:R58F	
  which	
  lead	
  to	
  
a	
  more	
  intense	
  iminium	
  species	
  (Figure	
   IV-­‐7).	
   As	
  it	
  was	
  mentioned	
  earlier	
  in	
  this	
  chapter,	
  our	
  
overreaching	
  goal	
  is	
  to	
  develop	
  fluorescent	
  proteins	
  that	
  span	
  the	
  entire	
  visible	
  spectra.	
  By	
  now	
  

140	
  

in	
  this	
  project,	
  the	
  absorption	
  maxima	
  are	
  tuned	
  from	
  501nm	
  to	
  705nm	
  and	
  emission	
  maxima	
  
from	
  613	
  nm	
  to	
  744	
  nm.	
  Covering	
  both	
  the	
  red	
  and	
  far-­‐red	
  fluorescence	
  wavelength	
  regimes.	
  	
  
This	
  fluorogenic	
  system	
  even	
  was	
  tested	
  in	
  mammalian	
  cells	
  for	
  live	
  cell	
  imaging	
  by	
  Dr.	
  
Santos.(111)	
   The	
   data	
   shows	
   that	
   ThioFluor	
   is	
   cell	
   permeable	
   and	
   displays	
   no	
   background	
  
fluorescence	
   when	
   excited	
   in	
   the	
   red,	
   even	
   when	
   long	
   incubation	
   periods	
   with	
   ThioFluor	
   are	
  
required,	
  unlike	
  merocyanine	
  aldehyde.	
  (Data	
  not	
  shown)	
  

IV-­‐6	
  Material	
  and	
  Methods	
  
Designed	
   hCRBPII	
   mutants	
   were	
   obtained	
   Dr.	
   Santos.	
   Mutants	
   were	
   expressed	
   and	
  
purified	
   according	
   to	
   the	
   protocol	
   described	
   in	
   Chapter	
   I.	
   The	
   purified	
   monomers	
   were	
   then	
  
used	
  for	
  crystallization	
  trials.	
  
Each	
   mutant	
   protein	
   was	
   concentrated	
   to	
   8mg/mL	
   in	
   a	
   buffer	
   containing	
   10	
   mM	
   Tris,	
  
100	
  mM	
  NaCl,	
  pH	
  =	
  8.0.	
  Proteins	
  were	
  incubated	
  with	
  4	
  equivalent	
  of	
  ThioFluor	
  ligand	
  (stock	
  
concentration	
  of	
  0.014M	
  in	
  DMSO)	
  in	
  room	
  temperature	
  overnight.	
  Crystals	
  were	
  grown	
  using	
  
the	
   hanging	
   drop	
   vapor	
   diffusion	
   method	
   using	
   1	
   ΟL	
   of	
   protein	
   solution	
   and	
   1	
   ΟL	
   of	
  
crystallization	
  solution	
  in	
  the	
  drop.	
  The	
  best	
  crystals	
  grew	
  using	
  a	
  crystallization	
  solution	
  of	
  25-­‐
40%	
   PEG4000,	
   0.1	
   M	
   NaCH3CO2Na	
   pH	
   4.5	
   0.1	
   M	
   NH4CH3CO2.	
   The	
   crystals	
   were	
   soaked	
   in	
   a	
  
cryoprotectant	
  solution	
  (same	
  crystallization	
  solution	
  with	
  15%	
  glycerol),	
  flash	
  frozen	
  in	
  loops	
  in	
  
liquid	
  nitrogen	
  and	
  stored	
  in	
  a	
  liquid	
  nitrogen	
  Dewar	
  prior	
  to	
  data	
  collection.	
  
Data	
  

collection	
  

and	
  

solving	
  

the	
  

structures	
  

for	
  

the	
  

crystallized	
  

mutants	
  

(Q108K:K40L:T53A:R58F,	
  Q108K:K40L:T51V:T53S:R58Y,	
  	
  Q108K:K40L:T51V:T53S:R58W:Y19W	
  
and	
  Q108K:K40L:T51V:T53S:R58W:Y19W:L117E,	
  Table	
   IV-­‐2)	
  	
  were	
  performed	
  according	
  to	
  the	
  

141	
  

method	
   described	
   in	
   Chapter	
   I.	
   The	
   chromophore	
   using	
   the	
   JLIGAND	
   program	
   to	
   generate	
  
restraints.(104,	
  105)	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

142	
  

Table	
  IV-­‐2:	
  X-­‐ray	
  crystallographic	
  data	
  and	
  refinement	
  statistics	
  for	
  monomeric	
  hCRBPII	
  
!
mutants.	
  
!

!
Q108K:K40L:T53A:R58F!

Q108K:K40L:T51A:!
T53S:R58Y!

	
  

P1!
36.51!

b(Å)!

55.16!

35.25!

36.43!

36.39!

c!(Å)!
ι(°!)!
	
   β(°!)!
δ(°!)!
Molecules!per!
Asymmetric!Unit!
Total!reflection!
Unique!Reflection!
Completeness!(%)!

68.87!
109.94!
98.49!
101.41!
4!

64.87!
92.51!
90.42!
111.72!
2!

64.08!
89.74!
89.36!
66.38!
2!

64.27!
89.84!
89.66!
66.15!
2!

597246!
82805!
95.71(91.0)!a!

268817!
47165!
79.63(94.6)a!

544327!
72294!
93.53(90.0)a!

623264!
69070!
94.8(89.0)a!

	
  

Average!I/σ!
Resolution!(Å)!(Last!Shell)!
Rwork/!Rfree!(%)!
RMSD!from!ideal!value!

!a

32.9(2.01) !
27.4X1.47(1.51X1.41)a!
23.3/26.3!
!

Bond!Length!(Å)!
0.008!
Bond!Angle!(°)!
1.249!
Average!B!factor!
15.89!
Number!of!water!
513!
molecules!
a!
Values!in!the!parenthesis!refer!to!the!last!resolution!shell.!

P1!
29.32!

Q108K:K40L:T51A:!
T53S:R58W:Y19W:L117E!

Space!group!
a(Å)!

	
  

P1!
32.29!

Q108K:K40L:T51A:!
T53S:R58W:Y19W!

a

43.37(1.64) !
29.2X1.34(1.37X1.34)a!

P1!
29.40!

a

44.63(6.9) !
29.60X1.220(1.25X1.29)a!

	
  

17.3/18.5!

0.013!
1.56!
20.58!
284!

0.01!
1.58!
12.07!
321!

0.007!
1.30!
13.66!
211!

143	
  

	
  

a

17.9/18.3!

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  

31.4(1.47) !
29.55X1.24(1.27X1.23)a!

23.8/28.2!
!

	
  

	
  

	
  
	
  

	
  

	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

REFERENCES	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  

144	
  

REFERENCES	
  
	
  
	
  
	
  
1.	
  
	
  
2.	
  
	
  
3.	
  

	
  
4.	
  
	
  
5.	
  
	
  
6.	
  

	
  
7.	
  
	
  
8.	
  
	
  
9.	
  
	
  
10.	
  
	
  
11.	
  

Ong,	
  D.	
  E.	
  (1984)	
  A	
  NOVEL	
  RETINOL-­‐BINDING	
  PROTEIN	
  FROM	
  RAT	
  -­‐	
  PURIFICATION	
  AND	
  
PARTIAL	
  CHARACTERIZATION,	
  Journal	
  of	
  Biological	
  Chemistry	
  259,	
  1476-­‐1482.	
  
Harder,	
  M.	
  E.,	
  Deinzer,	
  M.	
  L.,	
  Leid,	
  M.	
  E.,	
  and	
  Schimerlik,	
  M.	
  I.	
  (2004)	
  Global	
  analysis	
  of	
  
three-­‐state	
  protein	
  unfolding	
  data,	
  Protein	
  Science	
  13,	
  2207-­‐2222.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  crystallographica.	
  Section	
  
D,	
  Biological	
  crystallography	
  70,	
  3226-­‐3232.	
  
Veerkamp,	
   J.	
   H.,	
   Vankuppevelt,	
   T.,	
   Maatman,	
   R.,	
   and	
   Prinsen,	
   C.	
   F.	
   M.	
   (1993)	
  
STRUCTURAL	
   AND	
   FUNCTIONAL-­‐ASPECTS	
   OF	
   CYTOSOLIC	
   FATTY-­‐ACID-­‐BINDING	
  
PROTEINS,	
  Prostaglandins	
  Leukotrienes	
  and	
  Essential	
  Fatty	
  Acids	
  49,	
  887-­‐906.	
  
Banaszak,	
   L.,	
   Winter,	
   N.,	
   Xu,	
   Z.	
   H.,	
   Bernlohr,	
   D.	
   A.,	
   Cowan,	
   S.,	
   and	
   Jones,	
   T.	
   A.	
   (1994)	
  
LIPID-­‐BINDING	
   PROTEINS	
   -­‐	
   A	
   FAMILY	
   OF	
   FATTY-­‐ACID	
   AND	
   RETINOID	
   TRANSPORT	
  
PROTEINS,	
  Advances	
  in	
  Protein	
  Chemistry,	
  Vol	
  45	
  45,	
  89-­‐151.	
  
Vaezeslami,	
   S.,	
   Mathes,	
   E.,	
   Vasilelou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2006)	
   The	
  
structure	
   of	
   apo-­‐wild-­‐type	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   at	
   1.4	
   A	
   and	
   its	
  
relationship	
   to	
   ligand	
   binding	
   and	
   nuclear	
   translocation,	
   Journal	
   of	
   Molecular	
   Biology	
  
363,	
  687-­‐701.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins-­‐-­‐insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  lipid	
  research	
  43,	
  328-­‐349.	
  
Lucke,	
   C.,	
   Huang,	
   S.,	
   Rademacher,	
   M.,	
   and	
   Ruterjans,	
   H.	
   (2002)	
   New	
   insights	
   into	
  
intracellular	
   lipid	
   binding	
   proteins:	
   The	
   role	
   of	
   buried	
   water,	
   Protein	
   science	
   :	
   a	
  
publication	
  of	
  the	
  Protein	
  Society	
  11,	
  2382-­‐2392.	
  
Schaap,	
   F.	
   G.,	
   van	
   der	
   Vusse,	
   G.	
   J.,	
   and	
   Glatz,	
   J.	
   F.	
   (2002)	
   Evolution	
   of	
   the	
   family	
   of	
  
intracellular	
   lipid	
   binding	
   proteins	
   in	
   vertebrates,	
   Molecular	
   and	
   cellular	
   biochemistry	
  
239,	
  69-­‐77.	
  
Smathers,	
   R.	
   L.,	
   and	
   Petersen,	
   D.	
   R.	
   (2011)	
   The	
   human	
   fatty	
   acid-­‐binding	
   protein	
   family:	
  
evolutionary	
  divergences	
  and	
  functions,	
  Human	
  genomics	
  5,	
  170-­‐191.	
  
Franzoni,	
   L.,	
   Lucke,	
   C.,	
   Perez,	
   C.,	
   Cavazzini,	
   D.,	
   Rademacher,	
   M.,	
   Ludwig,	
   C.,	
   Spisni,	
   A.,	
  
Rossi,	
  G.	
  L.,	
  and	
  Ruterjans,	
  H.	
  (2002)	
  Structure	
  and	
  backbone	
  dynamics	
  of	
  apo-­‐	
  and	
  holo-­‐
145	
  

	
  
12.	
  

	
  
13.	
  

	
  
14.	
  

	
  
15.	
  
	
  
16.	
  
	
  
17.	
  

	
  
18.	
  
	
  
19.	
  
	
  
20.	
  
	
  
21.	
  

cellular	
   retinol-­‐binding	
   protein	
   in	
   solution,	
   Journal	
   of	
   Biological	
   Chemistry	
   277,	
   21983-­‐
21997.	
  
Crow,	
   J.	
   A.,	
   and	
   Ong,	
   D.	
   E.	
   (1985)	
   CELL-­‐SPECIFIC	
   IMMUNOHISTOCHEMICAL	
  
LOCALIZATION	
   OF	
   A	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   (TYPE-­‐2)	
   IN	
   THE	
   SMALL-­‐
INTESTINE	
  OF	
  RAT,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  
of	
  America	
  82,	
  4707-­‐4711.	
  
Vogel,	
  S.,	
  Mendelsohn,	
  C.	
  L.,	
  Mertz,	
  J.	
  R.,	
  Piantedosi,	
  R.,	
  Waldburger,	
  C.,	
  Gottesman,	
  M.	
  
E.,	
  and	
  Blaner,	
  W.	
  S.	
  (2001)	
  Characterization	
  of	
  a	
  new	
  member	
  of	
  the	
  fatty	
  acid-­‐binding	
  
protein	
   family	
   that	
   binds	
   all-­‐trans-­‐retinol,	
   Journal	
   of	
   Biological	
   Chemistry	
   276,	
   1353-­‐
1360.	
  
Folli,	
   C.,	
   Calderone,	
   V.,	
   Ottonello,	
   S.,	
   Bolchi,	
   A.,	
   Zanotti,	
   G.,	
   Stoppini,	
   M.,	
   and	
   Berni,	
   R.	
  
(2001)	
   Identification,	
   retinoid	
   binding,	
   and	
   x-­‐ray	
   analysis	
   of	
   a	
   human	
   retinol-­‐binding	
  
protein,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  98,	
  3710-­‐3715.	
  
Li,	
  E.,	
  Demmer,	
  L.	
  A.,	
  Sweetser,	
  D.	
  A.,	
  Ong,	
  D.	
  E.,	
  and	
  Gordon,	
  J.	
  I.	
  (1986)	
  ISOLATION	
  OF	
  A	
  
CLONED	
   CDNA-­‐ENCODING	
   AN	
   ABUNDANT	
   INTESTINAL	
   RETINOL	
   BINDING-­‐PROTEIN,	
  
Clinical	
  Research	
  34,	
  A685-­‐A685.	
  
Li,	
   E.,	
   and	
   Norris,	
   A.	
   W.	
   (1996)	
   Structure/function	
   of	
   cytoplasmic	
   vitamin	
   A-­‐binding	
  
proteins,	
  Annual	
  Review	
  of	
  Nutrition	
  16,	
  205-­‐234.	
  
Li,	
   E.,	
   Qian,	
   S.	
   J.,	
   Winter,	
   N.	
   S.,	
   Davignon,	
   A.,	
   Levin,	
   M.	
   S.,	
   and	
   Gordon,	
   J.	
   I.	
   (1991)	
  
FLUORINE	
   NUCLEAR-­‐MAGNETIC-­‐RESONANCE	
   ANALYSIS	
   OF	
   THE	
   LIGAND-­‐BINDING	
  
PROPERTIES	
   OF	
   2	
   HOMOLOGOUS	
   RAT	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEINS	
  
EXPRESSED	
  IN	
  ESCHERICHIA-­‐COLI,	
  Journal	
  of	
  Biological	
  Chemistry	
  266,	
  3622-­‐3629.	
  
Calderone,	
   V.,	
   Folli,	
   C.,	
   Marchesani,	
   A.,	
   Berni,	
   R.,	
   and	
   Zanotti,	
   G.	
   (2002)	
   Identification	
  
and	
   structural	
   analysis	
   of	
   a	
   zebrafish	
   apo	
   and	
   holo	
   cellular	
   retinol-­‐binding	
   protein,	
  
Journal	
  of	
  Molecular	
  Biology	
  321,	
  527-­‐535.	
  
Tarter,	
  M.,	
  Capaldi,	
  S.,	
  Carrizo,	
  M.	
  E.,	
  Ambrosi,	
  E.,	
  Perduca,	
  M.,	
  and	
  Monaco,	
  H.	
  L.	
  (2008)	
  
Crystal	
   structure	
   of	
   human	
   cellular	
   retinol-­‐binding	
   protein	
   11	
   to	
   1.2	
   angstrom	
  
resolution,	
  Proteins-­‐Structure	
  Function	
  and	
  Bioinformatics	
  70,	
  1626-­‐1630.	
  
Winter,	
   N.	
   S.,	
   Bratt,	
   J.	
   M.,	
   and	
   Banaszak,	
   L.	
   J.	
   (1993)	
   CRYSTAL-­‐STRUCTURES	
   OF	
   HOLO	
  
AND	
   APO-­‐CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN-­‐II,	
   Journal	
   of	
   Molecular	
   Biology	
   230,	
  
1247-­‐1259.	
  
Kane,	
  M.	
  A.,	
  Bright,	
  F.	
  V.,	
  and	
  Napoli,	
  J.	
  L.	
  (2011)	
  Binding	
  affinities	
  of	
  CRBPI	
  and	
  CRBPII	
  
for	
  9-­‐cis-­‐retinoids,	
  Biochimica	
  Et	
  Biophysica	
  Acta-­‐General	
  Subjects	
  1810,	
  514-­‐518.	
  
146	
  

22.	
  
	
  
23.	
  

	
  
24.	
  
	
  
25.	
  
	
  
26.	
  
	
  
27.	
  

	
  
28.	
  
	
  
29.	
  
	
  
30.	
  
	
  
31.	
  
	
  
32.	
  

Malpeli,	
   G.,	
   Stoppini,	
   M.,	
   Zapponi,	
   M.	
   C.,	
   Folli,	
   C.,	
   and	
   Berni,	
   R.	
   (1995)	
   INTERACTIONS	
  
WITH	
   RETINOL	
   AND	
   RETINOIDS	
   OF	
   BOVINE	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN,	
  
European	
  Journal	
  of	
  Biochemistry	
  229,	
  486-­‐493.	
  
McBee,	
   J.	
   K.,	
   Kuksa,	
   V.,	
   Alvarez,	
   R.,	
   de	
   Lera,	
   A.	
   R.,	
   Prezhdo,	
   O.,	
   Haeseleer,	
   F.,	
   Sokal,	
   I.,	
  
and	
   Palczewski,	
   K.	
   (2000)	
   Isomerization	
   of	
   all-­‐trans-­‐retinol	
   to	
   cis-­‐retinols	
   in	
   bovine	
  
retinal	
   pigment	
   epithelial	
   cells:	
   Dependence	
   on	
   the	
   specificity	
   of	
   retinoid-­‐binding	
  
proteins,	
  Biochemistry	
  39,	
  11370-­‐11380.	
  
Borshchevskiy,	
  V.	
  I.,	
  Round,	
  E.	
  S.,	
  Popov,	
  A.	
  N.,	
  Bueldt,	
  G.,	
  and	
  Gordeliy,	
  V.	
  I.	
  (2011)	
  X-­‐
ray-­‐Radiation-­‐Induced	
   Changes	
   in	
   Bacteriorhodopsin	
   Structure,	
   Journal	
   of	
   Molecular	
  
Biology	
  409,	
  813-­‐825.	
  
Matsui,	
  Y.,	
  Sakai,	
  K.,	
  Murakami,	
  M.,	
  Shiro,	
  Y.,	
  Adachi,	
  S.,	
  Okumura,	
  H.,	
  and	
  Kouyama,	
  T.	
  
(2002)	
  Specific	
  damage	
  induced	
  by	
  X-­‐ray	
  radiation	
  and	
  structural	
  changes	
  in	
  the	
  primary	
  
photoreaction	
  of	
  bacteriorhodopsin,	
  Journal	
  of	
  Molecular	
  Biology	
  324,	
  469-­‐481.	
  
Schreckenbach,	
   T.,	
   Walckhoff,	
   B.,	
   and	
   Oesterhelt,	
   D.	
   (1977)	
   STUDIES	
   ON	
   RETINAL-­‐
PROTEIN	
  INTERACTION	
  IN	
  BACTERIORHODOPSIN,	
  European	
  Journal	
  of	
  Biochemistry	
  76,	
  
499-­‐511.	
  
Cowan,	
   S.	
   W.,	
   Newcomer,	
   M.	
   E.,	
   and	
   Jones,	
   T.	
   A.	
   (1993)	
   CRYSTALLOGRAPHIC	
   STUDIES	
  
ON	
   A	
   FAMILY	
   OF	
   CELLULAR	
   LIPOPHILIC	
   TRANSPORT	
   PROTEINS	
   -­‐	
   REFINEMENT	
   OF	
   P2	
  
MYELIN	
   PROTEIN	
   AND	
   THE	
   STRUCTURE	
   DETERMINATION	
   AND	
   REFINEMENT	
   OF	
  
CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   IN	
   COMPLEX	
   WITH	
   ALL-­‐TRANS-­‐RETINOL,	
   Journal	
  
of	
  Molecular	
  Biology	
  230,	
  1225-­‐1246.	
  
Vasileiou,	
   C.,	
   Vaezeslami,	
   S.,	
   Crist,	
   R.	
   M.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2007)	
   Protein	
   design:	
   reengineering	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   into	
   a	
  
rhodopsin	
  protein	
  mimic,	
  J	
  Am	
  Chem	
  Soc	
  129,	
  6140-­‐6148.	
  
Otwinowski,	
   Z.,	
   and	
   Minor,	
   W.	
   (1997)	
   Processing	
   of	
   X-­‐ray	
   diffraction	
   data	
   collected	
   in	
  
oscillation	
  mode,	
  Macromolecular	
  Crystallography,	
  Pt	
  A	
  276,	
  307-­‐326.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (1997)	
   MOLREP:	
   an	
   automated	
   program	
   for	
   molecular	
  
replacement,	
  Journal	
  of	
  Applied	
  Crystallography	
  30,	
  1022-­‐1025.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (2010)	
   Molecular	
   replacement	
   with	
   MOLREP,	
   Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  22-­‐25.	
  
Murshudov,	
   G.	
   N.,	
   Skubak,	
   P.,	
   Lebedev,	
   A.	
   A.,	
   Pannu,	
   N.	
   S.,	
   Steiner,	
   R.	
   A.,	
   Nicholls,	
   R.	
   A.,	
  
Winn,	
   M.	
   D.,	
   Long,	
   F.,	
   and	
   Vagin,	
   A.	
   A.	
   (2011)	
   REFMAC5	
   for	
   the	
   refinement	
   of	
  
macromolecular	
   crystal	
   structures,	
   Acta	
   Crystallographica	
   Section	
   D-­‐Biological	
  
Crystallography	
  67,	
  355-­‐367.	
  
147	
  

33.	
  
	
  
34.	
  
	
  
35.	
  
	
  
36.	
  
	
  
37.	
  
	
  
38.	
  
	
  
39.	
  
	
  
40.	
  
	
  
41.	
  
	
  
42.	
  
	
  
43.	
  
	
  
44.	
  
	
  
45.	
  

Emsley,	
   P.,	
   Lohkamp,	
   B.,	
   Scott,	
   W.	
   G.,	
   and	
   Cowtan,	
   K.	
   (2010)	
   Features	
   and	
   development	
  
of	
  Coot,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  486-­‐501.	
  
Bennett,	
   M.	
   J.,	
   Schlunegger,	
   M.	
   P.,	
   and	
   Eisenberg,	
   D.	
   (1995)	
   3D	
   domain	
   swapping	
   -­‐	
   A	
  
mechanism	
  for	
  oligomer	
  assembly,	
  Protein	
  Science	
  4,	
  2455-­‐2468.	
  
Crestfield,	
   A.	
   M.,	
   Stein,	
   W.	
   H.,	
   and	
   Moore,	
   S.	
   (1962)	
   On	
   the	
   aggregation	
   of	
   bovine	
  
pancreatic	
  ribonuclease,	
  Archives	
  of	
  biochemistry	
  and	
  biophysics	
  Suppl	
  1,	
  217-­‐222.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   As	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  Science	
  11,	
  1285-­‐1299.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   Structures	
   of	
   the	
   two	
   3D	
  
domain-­‐swapped	
  RNase	
  A	
  trimers,	
  Protein	
  Science	
  11,	
  371-­‐380.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   S.	
   (2001)	
   Crystal	
   structure	
   of	
   an	
  
Rnase	
  A	
  dimer	
  displaying	
  a	
  new	
  type	
  of	
  3D	
  domain	
  swapping,	
  Nucleic	
  Acid	
  Database.	
  
Liu,	
  Y.	
  S.,	
  Hart,	
  P.	
  J.,	
  Schlunegger,	
  M.	
  P.,	
  and	
  Eisenberg,	
  D.	
  (1998)	
  The	
  crystal	
  structure	
  of	
  
a	
   3D	
   domain-­‐swapped	
   dimer	
   of	
   RNase	
   A	
   at	
   a	
   2.1-­‐angstrom	
   resolution,	
   Proceedings	
   of	
  
the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  95,	
  3437-­‐3442.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   as	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  11,	
  1285-­‐1299.	
  
Koharudin,	
   L.	
   M.,	
   Liu,	
   L.,	
   and	
   Gronenborn,	
   A.	
   M.	
   (2013)	
   Different	
   3D	
   domain-­‐swapped	
  
oligomeric	
   cyanovirin-­‐N	
   structures	
   suggest	
   trapped	
   folding	
   intermediates,	
   Proc	
   Natl	
  
Acad	
  Sci	
  U	
  S	
  A	
  110,	
  7702-­‐7707.	
  
Assar,	
  Z.,	
  Nossoni,	
  Z.,	
  Wang,	
  W.,	
  Santos,	
  E.	
  M.,	
  Kramer,	
  K.,	
  McCornack,	
  C.,	
  Vasileiou,	
  C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   Domain-­‐Swapped	
   Dimers	
   of	
   Intracellular	
   Lipid-­‐
Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure	
  24,	
  1590-­‐1598.	
  
Bax,	
  B.,	
  Lapatto,	
  R.,	
  Nalini,	
  V.,	
  Driessen,	
  H.,	
  Lindley,	
  P.	
  F.,	
  Mahadevan,	
  D.,	
  Blundell,	
  T.	
  L.,	
  
and	
   Slingsby,	
   C.	
   (1990)	
   X-­‐ray	
   analysis	
   of	
   beta-­‐B2-­‐crystallin	
   and	
   evolution	
   of	
   oligomeric	
  
lens	
  proteins,	
  Nature	
  347,	
  776-­‐780.	
  
Hakansson,	
  M.,	
  Svensson,	
  A.,	
  Fast,	
  J.,	
  and	
  Linse,	
  S.	
  (2001)	
  An	
  extended	
  hydrophobic	
  core	
  
induces	
  EF-­‐hand	
  swapping,	
  Protein	
  Science	
  10,	
  927-­‐933.	
  
Rousseau,	
   F.,	
   Schymkowitz,	
   J.,	
   and	
   Itzhaki,	
   L.	
   S.	
   (2012)	
   Implications	
   of	
   3D	
   domain	
  
swapping	
   for	
   protein	
   folding,	
   misfolding	
   and	
   function,	
   In	
   Protein	
   Dimerization	
   and	
  
Oligomerization	
  in	
  Biology	
  (Mattews,	
  J.	
  M.,	
  Ed.),	
  pp	
  137-­‐152.	
  

	
  
148	
  

46.	
  

	
  
47.	
  
	
  
48.	
  
	
  
49.	
  
	
  
50.	
  
	
  
51.	
  
	
  
52.	
  
	
  
53.	
  
	
  
54.	
  
	
  
55.	
  
	
  
56.	
  
	
  

Yang,	
   F.,	
   Bewley,	
   C.	
   A.,	
   Louis,	
   J.	
   M.,	
   Gustafson,	
   K.	
   R.,	
   Boyd,	
   M.	
   R.,	
   Gronenborn,	
   A.	
   M.,	
  
Clore,	
  G.	
  M.,	
  and	
  Wlodawer,	
  A.	
  (1999)	
  Crystal	
  structure	
  of	
  cyanovirin-­‐N,	
  a	
  potent	
  HIV-­‐
inactivating	
  protein,	
  shows	
  unexpected	
  domain	
  swapping,	
  Journal	
  of	
  Molecular	
  Biology	
  
288,	
  403-­‐412.	
  
Szilagyi,	
   A.,	
   Zhang,	
   Y.,	
   and	
   Zavodszky,	
   P.	
   (2012)	
   Intra-­‐Chain	
   3D	
   Segment	
   Swapping	
  
Spawns	
  the	
  Evolution	
  of	
  New	
  Multidomain	
  Protein	
  Architectures,	
  Journal	
  of	
  Molecular	
  
Biology	
  415,	
  221-­‐235.	
  
Stroud,	
   J.	
   C.,	
   Liu,	
   C.,	
   Teng,	
   P.	
   K.,	
   and	
   Eisenberg,	
   D.	
   (2012)	
   Toxic	
   fibrillar	
   oligomers	
   of	
  
amyloid-­‐beta	
   have	
   cross-­‐beta	
   structure,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  109,	
  7717-­‐7722.	
  
Cleary,	
   J.	
   P.,	
   Walsh,	
   D.	
   M.,	
   Hofmeister,	
   J.	
   J.,	
   Shankar,	
   G.	
   M.,	
   Kuskowski,	
   M.	
   A.,	
   Selkoe,	
   D.	
  
J.,	
   and	
   Ashe,	
   K.	
   H.	
   (2005)	
   Natural	
   oligomers	
   of	
   the	
   amyloid-­‐protein	
   specifically	
   disrupt	
  
cognitive	
  function,	
  Nature	
  Neuroscience	
  8,	
  79-­‐84.	
  
Fandrich,	
   M.,	
   Fletcher,	
   M.	
   A.,	
   and	
   Dobson,	
   C.	
   M.	
   (2001)	
   Amyloid	
   fibrils	
   from	
   muscle	
  
myoglobin	
   -­‐	
   Even	
   an	
   ordinary	
   globular	
   protein	
   can	
   assume	
   a	
   rogue	
   guise	
   if	
   conditions	
  
are	
  right,	
  Nature	
  410,	
  165-­‐166.	
  
Gronenborn,	
   A.	
   M.	
   (2009)	
   Protein	
   acrobatics	
   in	
   pairs	
   -­‐	
   dimerization	
   via	
   domain	
  
swapping,	
  Current	
  Opinion	
  in	
  Structural	
  Biology	
  19,	
  39-­‐49.	
  
Schlunegger,	
  M.	
  P.,	
  Bennett,	
  M.	
  J.,	
  and	
  Eisenberg,	
  D.	
  (1997)	
  Oligomer	
  formation	
  by	
  3D	
  
domain	
   swapping:	
   A	
   model	
   for	
   protein	
   assembly	
   and	
   misassembly,	
   In	
   Advances	
   in	
  
Protein	
  Chemistry,	
  Vol	
  50:	
  Protein	
  Misassembly	
  (Wetzel,	
  R.,	
  Ed.),	
  pp	
  61-­‐122.	
  
Selkoe,	
  D.	
  J.	
  (2004)	
  Folding	
  proteins	
  in	
  fatal	
  ways	
  (vol	
  426,	
  pg	
  900,	
  2003),	
  Nature	
  428,	
  
445-­‐445.	
  
Shtilerman,	
  M.	
  D.,	
  Ding,	
  T.	
  T.,	
  and	
  Lansbury,	
  P.	
  T.	
  (2002)	
  Molecular	
  crowding	
  accelerates	
  
fibrillization	
   of	
   alpha-­‐synuclein:	
   Could	
   an	
   increase	
   in	
   the	
   cytoplasmic	
   protein	
  
concentration	
  induce	
  Parkinson's	
  disease?,	
  Biochemistry	
  41,	
  3855-­‐3860.	
  
Sunde,	
  M.,	
  and	
  Blake,	
  C.	
  (1997)	
  The	
  structure	
  of	
  amyloid	
  fibrils	
  by	
  electron	
  microscopy	
  
and	
   X-­‐ray	
   diffraction,	
   In	
   Advances	
   in	
   Protein	
   Chemistry,	
   Vol	
   50:	
   Protein	
   Misassembly	
  
(Wetzel,	
  R.,	
  Ed.),	
  pp	
  123-­‐159.	
  
Bennett,	
  M.	
  J.,	
  Choe,	
  S.,	
  and	
  Eisenberg,	
  D.	
  (1994)	
  Refined	
  structure	
  of	
  dimeric	
  diphtheria	
  
toxin	
  at	
  2.0	
  A	
  resolution,	
  Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  3,	
  1444-­‐
1463.	
  

149	
  

57.	
  
	
  
58.	
  
	
  
59.	
  
	
  
60.	
  
	
  
61.	
  
	
  
62.	
  
	
  
63.	
  
	
  
64.	
  
	
  
65.	
  
	
  
66.	
  
	
  
67.	
  
	
  
68.	
  
	
  
69.	
  
	
  

O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
  environment,	
  Curr	
  Opin	
  Struct	
  Biol	
  11,	
  114-­‐119.	
  
Bennett,	
   M.	
   J.,	
   Choe,	
   S.,	
   and	
   Eisenberg,	
   D.	
   (1994)	
   Rlefined	
   strucutre	
   of	
   dimeric	
  
diphtheria-­‐toxin	
  at	
  2.0-­‐angstrom	
  resolution,	
  Protein	
  Science	
  3,	
  1444-­‐1463.	
  
Carroll,	
   S.	
   F.,	
   Barbieri,	
   J.	
   T.,	
   and	
   Collier,	
   R.	
   J.	
   (1986)	
   Dimeric	
   form	
   of	
   diphtheria-­‐toxin-­‐
purification	
  and	
  characterization,	
  Biochemistry	
  25,	
  2425-­‐2430.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
   environment	
   (vol	
   11,	
   pg	
   114,	
   2001),	
   Current	
   Opinion	
   in	
   Structural	
   Biology	
  
11,	
  500-­‐500.	
  
Minton,	
   A.	
   P.	
   (2000)	
   Implications	
   of	
   macromolecular	
   crowding	
   for	
   protein	
   assembly,	
  
Current	
  Opinion	
  in	
  Structural	
  Biology	
  10,	
  34-­‐39.	
  
O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   J.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nature	
  chemical	
  biology	
  10,	
  884-­‐891.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2004)	
  Monitoring	
  protein	
  stability	
  and	
  aggregation	
  in	
  
vivo	
  by	
  real-­‐time	
  fluorescent	
  labeling,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  101,	
  523-­‐528.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2006)	
  Inhibition	
  of	
  protein	
  aggregation	
  in	
  vitro	
  and	
  in	
  
vivo	
  by	
  a	
  natural	
  osmoprotectant,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  103,	
  13357-­‐13361.	
  
Ropson,	
  I.	
  J.,	
  Boyer,	
  J.	
  A.,	
  Schaeffer,	
  B.	
  A.,	
  and	
  Dalessio,	
  P.	
  M.	
  (2009)	
  Comparison	
  of	
  the	
  
folding	
   mechanism	
   of	
   highly	
   homologous	
   proteins	
   in	
   the	
   lipid-­‐binding	
   protein	
   family,	
  
Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  75,	
  799-­‐806.	
  
Burns,	
   L.	
   L.,	
   and	
   Ropson,	
   I.	
   J.	
   (2001)	
   Folding	
   of	
   intracellular	
   retinol	
   and	
   retinoic	
   acid	
  
binding	
  proteins,	
  Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  43,	
  292-­‐302.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nat.	
  Chem.	
  Biol.	
  10,	
  884-­‐891.	
  

150	
  

70.	
  
	
  
71.	
  
	
  
72.	
  
	
  
73.	
  
	
  
74.	
  
	
  
75.	
  
	
  
76.	
  
	
  
77.	
  
	
  
78.	
  
	
  
79.	
  

	
  
80.	
  

Crist,	
   R.	
   M.,	
   Vasileiou,	
   C.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2006)	
  
Engineering	
  a	
  Rhodopsin	
  Protein	
  Mimic,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  128,	
  
4522-­‐4523.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  J	
  Am	
  Chem	
  Soc	
  135,	
  16111-­‐16119.	
  
Wang,	
  W.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2014)	
  The	
  photochemical	
  determinants	
  of	
  color	
  
vision:	
   revealing	
   how	
   opsins	
   tune	
   their	
   chromophore's	
   absorption	
   wavelength,	
  
BioEssays	
  :	
  news	
  and	
  reviews	
  in	
  molecular,	
  cellular	
  and	
  developmental	
  biology	
  36,	
  65-­‐74.	
  
Wang,	
   W.,	
   Nossoni,	
   Z.,	
   Berbasova,	
   T.,	
   Watson,	
   C.	
   T.,	
   Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Vasileiou,	
   C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   electronic	
   absorption	
   of	
   protein-­‐
embedded	
  all-­‐trans-­‐retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
Nosrati,	
   M.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   A	
  
Photoisomerizing	
  Rhodopsin	
  Mimic	
  Observed	
  at	
  Atomic	
  Resolution,	
  J	
  Am	
  Chem	
  Soc	
  138,	
  
8802-­‐8808.	
  
Berbasova,	
   T.,	
   Santos,	
   E.	
   M.,	
   Nosrati,	
   M.,	
   Vasileiou,	
   C.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2016)	
  Light-­‐Activated	
  Reversible	
  Imine	
  Isomerization:	
  Towards	
  a	
  Photochromic	
  Protein	
  
Switch,	
  Chembiochem	
  :	
  a	
  European	
  journal	
  of	
  chemical	
  biology	
  17,	
  407-­‐414.	
  
Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Berbasova,	
   T.,	
   Nosrati,	
   M.,	
   Jia,	
   X.,	
   Vasileiou,	
   C.,	
   Wang,	
   W.,	
   Santos,	
   E.	
  
M.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2015)	
  "Turn-­‐on"	
  protein	
  fluorescence:	
  in	
  situ	
  formation	
  
of	
  cyanine	
  dyes,	
  J	
  Am	
  Chem	
  Soc	
  137,	
  1073-­‐1080.	
  
Wang,	
  W.,	
  Nossoni,	
  Z.,	
  Berbasova,	
  T.,	
  Watson,	
  C.	
  T.,	
  Yapici,	
  I.,	
  Lee,	
  K.	
  S.	
  S.,	
  Vasileiou,	
  C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   Electronic	
   Absorption	
   of	
   Protein-­‐
Embedded	
  All-­‐trans-­‐Retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
Gunasekaran,	
   K.,	
   Hagler,	
   A.	
   T.,	
   and	
   Gierasch,	
   L.	
   M.	
   (2004)	
   Sequence	
   and	
   structural	
  
analysis	
   of	
   cellular	
   retinoic	
   acid-­‐binding	
   proteins	
   reveals	
   a	
   network	
   of	
   conserved	
  
hydrophobic	
  interactions,	
  Proteins-­‐Structure	
  Function	
  and	
  Genetics	
  54,	
  179-­‐194.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  Crystallogr.	
  Sect.	
  D.	
  Biol.	
  
Crystallogr.	
  70,	
  3226-­‐3232.	
  
Shameer,	
   K.,	
   Shingate,	
   P.	
   N.,	
   Manjunath,	
   S.	
   C.	
   P.,	
   Karthika,	
   M.,	
   Pugalenthi,	
   G.,	
   and	
  
Sowdhamini,	
   R.	
   (2011)	
   3DSwap:	
   curated	
   knowledgebase	
   of	
   proteins	
   involved	
   in	
   3D	
  
domain	
  swapping,	
  Database-­‐the	
  Journal	
  of	
  Biological	
  Databases	
  and	
  Curation.	
  
151	
  

81.	
  

	
  
82.	
  
	
  
83.	
  
	
  
84.	
  
	
  
85.	
  
	
  
86.	
  
	
  
87.	
  
	
  
88.	
  
	
  
89.	
  
	
  
90.	
  
	
  
91.	
  
	
  
92.	
  

Sillitoe,	
   I.,	
   Cuff,	
   A.	
   L.,	
   Dessailly,	
   B.	
   H.,	
   Dawson,	
   N.	
   L.,	
   Furnham,	
   N.,	
   Lee,	
   D.,	
   Lees,	
   J.	
   G.,	
  
Lewis,	
   T.	
   E.,	
   Studer,	
   R.	
   A.,	
   Rentzsch,	
   R.,	
   Yeats,	
   C.,	
   Thornton,	
   J.	
   M.,	
   and	
   Orengo,	
   C.	
   A.	
  
(2013)	
  New	
  functional	
  families	
  (FunFams)	
  in	
  CATH	
  to	
  improve	
  the	
  mapping	
  of	
  conserved	
  
functional	
  sites	
  to	
  3D	
  structures,	
  Nucleic	
  Acids	
  Research	
  41,	
  D490-­‐D498.	
  
Englander,	
   S.	
   W.,	
   and	
   Mayne,	
   L.	
   (2014)	
   The	
   nature	
   of	
   protein	
   folding	
   pathways,	
   Proc	
  
Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  111,	
  15873-­‐15880.	
  
Englander,	
  S.	
  W.,	
  Mayne,	
  L.,	
  and	
  Krishna,	
  M.	
  M.	
  (2007)	
  Protein	
  folding	
  and	
  misfolding:	
  
mechanism	
  and	
  principles,	
  Quarterly	
  reviews	
  of	
  biophysics	
  40,	
  287-­‐326.	
  
Krishna,	
   M.	
   M.,	
   and	
   Englander,	
   S.	
   W.	
   (2007)	
   A	
   unified	
   mechanism	
   for	
   protein	
   folding:	
  
predetermined	
   pathways	
   with	
   optional	
   errors,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  16,	
  449-­‐464.	
  
Mayne,	
   L.,	
   and	
   Englander,	
   S.	
   W.	
   (2000)	
   Two-­‐state	
   vs.	
   multistate	
   protein	
   unfolding	
  
studied	
   by	
   optical	
   melting	
   and	
   hydrogen	
   exchange,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  9,	
  1873-­‐1877.	
  
Roder,	
  H.,	
  Elove,	
  G.	
  A.,	
  and	
  Englander,	
  S.	
  W.	
  (1988)	
  Structural	
  characterization	
  of	
  folding	
  
intermediates	
   in	
   cytochrome	
   c	
   by	
   H-­‐exchange	
   labelling	
   and	
   proton	
   NMR,	
   Nature	
   335,	
  
700-­‐704.	
  
Martins,	
   N.	
   F.,	
   and	
   Santoro,	
   M.	
   M.	
   (1999)	
   Partially	
   folded	
   intermediates	
   during	
  
trypsinogen	
  denaturation,	
  Brazilian	
  journal	
  of	
  medical	
  and	
  biological	
  research	
  =	
  Revista	
  
brasileira	
  de	
  pesquisas	
  medicas	
  e	
  biologicas	
  32,	
  673-­‐682.	
  
Medina,	
   E.,	
   Cordova,	
   C.,	
   Villalobos,	
   P.,	
   Reyes,	
   J.,	
   Komives,	
   E.	
   A.,	
   Ramirez-­‐Sarmiento,	
   C.	
  
A.,	
   and	
   Babul,	
   J.	
   (2016)	
   Three-­‐Dimensional	
   Domain	
   Swapping	
   Changes	
   the	
   Folding	
  
Mechanism	
  of	
  the	
  Forkhead	
  Domain	
  of	
  FoxP1,	
  Biophysical	
  journal	
  110,	
  2349-­‐2360.	
  
Zwanzig,	
   R.	
   (1997)	
   Two-­‐state	
   models	
   of	
   protein	
   folding	
   kinetics,	
   Proceedings	
   of	
   the	
  
National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  94,	
  148-­‐150.	
  
Sanson,	
   B.,	
   Wang,	
   T.,	
   Sun,	
   J.,	
   Wang,	
   L.,	
   Kaczocha,	
   M.,	
   Ojima,	
   I.,	
   Deutsch,	
   D.,	
   and	
   Li,	
   H.	
  
(2014)	
  Crystallographic	
  study	
  of	
  FABP5	
  as	
  an	
  intracellular	
  endocannabinoid	
  transporter,	
  
Acta	
  crystallographica.	
  Section	
  D,	
  Biological	
  crystallography	
  70,	
  290-­‐298.	
  
Assar,	
   Z.,	
   Nossoni,	
   Z.,	
   Wang,	
   W.,	
   Santos,	
   Elizabeth	
  M.,	
   Kramer,	
   K.,	
   McCornack,	
   C.,	
  
Vasileiou,	
  C.,	
  Borhan,	
  B.,	
  and	
  Geiger,	
  James	
  H.	
  Domain-­‐Swapped	
  Dimers	
  of	
  Intracellular	
  
Lipid-­‐Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure.	
  
Silvaroli,	
   J.	
   A.,	
   Arne,	
   J.	
   M.,	
   Chelstowska,	
   S.,	
   Kiser,	
   P.	
   D.,	
   Banerjee,	
   S.,	
   and	
   Golczak,	
   M.	
  
(2016)	
   Ligand	
   Binding	
   Induces	
   Conformational	
   Changes	
   in	
   Human	
   Cellular	
   Retinol-­‐
152	
  

	
  
93.	
  
	
  
94.	
  
	
  
95.	
  

	
  
96.	
  
	
  
97.	
  

	
  
98.	
  
	
  
99.	
  

	
  
100.	
  
	
  
101.	
  

	
  

binding	
  Protein	
  1	
  (CRBP1)	
  Revealed	
  by	
  Atomic	
  Resolution	
  Crystal	
  Structures,	
  The	
  Journal	
  
of	
  biological	
  chemistry	
  291,	
  8528-­‐8540.	
  
Gillilan,	
  R.	
  E.,	
  Ayers,	
  S.	
  D.,	
  and	
  Noy,	
  N.	
  (2007)	
  Structural	
  basis	
  for	
  activation	
  of	
  fatty	
  acid-­‐
binding	
  protein	
  4,	
  J	
  Mol	
  Biol	
  372,	
  1246-­‐1260.	
  
Sessler,	
  R.	
  J.,	
  and	
  Noy,	
  N.	
  (2005)	
  A	
  ligand-­‐activated	
  nuclear	
  localization	
  signal	
  in	
  cellular	
  
retinoic	
  acid	
  binding	
  protein-­‐II,	
  Molecular	
  cell	
  18,	
  343-­‐353.	
  
Armstrong,	
  E.	
  H.,	
  Goswami,	
  D.,	
  Griffin,	
  P.	
  R.,	
  Noy,	
  N.,	
  and	
  Ortlund,	
  E.	
  A.	
  (2014)	
  Structural	
  
Basis	
   for	
   Ligand	
   Regulation	
   of	
   the	
   Fatty	
   Acid-­‐binding	
   Protein	
   5,	
   Peroxisome	
   Proliferator-­‐
activated	
   Receptor	
   β/δ	
   (FABP5-­‐PPARβ/δ)	
   Signaling	
   Pathway,	
   The	
   Journal	
   of	
   biological	
  
chemistry	
  289,	
  14941-­‐14954.	
  
Greenfield,	
  N.	
  J.	
  (2006)	
  Using	
  circular	
  dichroism	
  collected	
  as	
  a	
  function	
  of	
  temperature	
  
to	
  determine	
  the	
  thermodynamics	
  of	
  protein	
  unfolding	
  and	
  binding	
  interactions,	
  Nature	
  
Protocols	
  1,	
  2527-­‐2535.	
  
Ramos,	
   C.	
   R.	
   R.,	
   Oyama,	
   S.,	
   Jr.,	
   Sforca,	
   M.	
   L.,	
   Pertinhez,	
   T.	
   A.,	
   Ho,	
   P.	
   L.,	
   and	
   Spisni,	
   A.	
  
(2008)	
   2POA:	
   Schistosoma	
   mansoni	
   Sm14	
   Fatty	
   Acid-­‐Binding	
   Protein:	
   improvement	
   of	
  
protein	
   stability	
   by	
   substitution	
   of	
   the	
   single	
   Cys62	
   residue,	
   Worldwide	
   Protein	
   Data	
  
Bank.	
  
Gill,	
   S.	
   C.,	
   and	
   Vonhippel,	
   P.	
   H.	
   (1989)	
   CALCULATION	
   OF	
   PROTEIN	
   EXTINCTION	
  
COEFFICIENTS	
   FROM	
   AMINO-­‐ACID	
   SEQUENCE	
   DATA,	
   Analytical	
   Biochemistry	
   182,	
   319-­‐
326.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Davis,	
  I.	
  W.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  
J.,	
   Hung,	
   L.	
   W.,	
   Kapral,	
   G.	
   J.,	
   Grosse-­‐Kunstleve,	
   R.	
   W.,	
   McCoy,	
   A.	
   J.,	
   Moriarty,	
   N.	
   W.,	
  
Oeffner,	
  R.,	
  Read,	
  R.	
  J.,	
  Richardson,	
  D.	
  C.,	
  Richardson,	
  J.	
  S.,	
  Terwilliger,	
  T.	
  C.,	
  and	
  Zwart,	
  
P.	
   H.	
   (2010)	
   PHENIX:	
   a	
   comprehensive	
   Python-­‐based	
   system	
   for	
   macromolecular	
  
structure	
  solution,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  213-­‐
221.	
  
Bailey,	
  S.	
  (1994)	
  THE	
  CCP4	
  SUITE	
  -­‐	
  PROGRAMS	
  FOR	
  PROTEIN	
  CRYSTALLOGRAPHY,	
  Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  50,	
  760-­‐763.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  J.,	
  Hung,	
  L.-­‐
W.,	
  Jain,	
  S.,	
  Kapral,	
  G.	
  J.,	
  Kunstleve,	
  R.	
  W.	
  G.,	
  McCoy,	
  A.	
  J.,	
  Moriarty,	
  N.	
  W.,	
  Oeffner,	
  R.	
  
D.,	
   Read,	
   R.	
   J.,	
   Richardson,	
   D.	
   C.,	
   Richardson,	
   J.	
   S.,	
   Terwilliger,	
   T.	
   C.,	
   and	
   Zwart,	
   P.	
   H.	
  
(2011)	
  The	
  Phenix	
  software	
  for	
  automated	
  determination	
  of	
  macromolecular	
  structures,	
  
Methods	
  55,	
  94-­‐106.	
  

153	
  

102.	
  
103.	
  
	
  
104.	
  
	
  
105.	
  
	
  
106.	
  
	
  
107.	
  

	
  
108.	
  

	
  
109.	
  
	
  
110.	
  
	
  
111.	
  
	
  
112.	
  

Murshudov,	
  G.	
  N.,	
  Vagin,	
  A.	
  A.,	
  and	
  Dodson,	
  E.	
  J.	
  (1997)	
  Refinement	
  of	
  macromolecular	
  
structures	
   by	
   the	
   maximum-­‐likelihood	
   method,	
   Acta	
   Crystallographica	
   Section	
   D-­‐
Biological	
  Crystallography	
  53,	
  240-­‐255.	
  
Ong,	
   D.	
   E.,	
   and	
   Page,	
   D.	
   L.	
   (1987)	
   Cellular	
   retinol-­‐binding	
   protein	
   (type	
   two)	
   is	
   abundant	
  
in	
  human	
  small	
  intestine,	
  Journal	
  of	
  Lipid	
  Research	
  28,	
  739-­‐745.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins	
   –	
   insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  Lipid	
  Research	
  43,	
  328-­‐349.	
  
Gillilan,	
   R.	
   E.,	
   Ayers,	
   S.	
   D.,	
   and	
   Noy,	
   N.	
   (2007)	
   Structural	
   Basis	
   for	
   Activation	
   of	
   Fatty	
  
Acid-­‐binding	
  Protein	
  4,	
  Journal	
  of	
  Molecular	
  Biology	
  372,	
  1246-­‐1260.	
  
Kaczocha,	
   M.,	
   Glaser,	
   S.	
   T.,	
   and	
   Deutsch,	
   D.	
   G.	
   (2009)	
   Identification	
   of	
   intracellular	
  
carriers	
  for	
  the	
  endocannabinoid	
  anandamide,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  106,	
  6375-­‐6380.	
  
Sulsky,	
  R.,	
  Magnin,	
  D.	
  R.,	
  Huanga,	
  Y.,	
  Simpkins,	
  L.,	
  Taunk,	
  P.,	
  Patel,	
  M.,	
  Zhu,	
  Y.,	
  Stouch,	
  T.	
  
R.,	
  Bassolino-­‐Klimas,	
  D.,	
  Parker,	
  R.,	
  Harrity,	
  T.,	
  Stoffel,	
  R.,	
  Taylor,	
  D.	
  S.,	
  Lavoie,	
  T.	
  B.,	
  Kish,	
  
K.,	
  Jacobson,	
  B.	
  L.,	
  Sheriff,	
  S.,	
  Adam,	
  L.	
  P.,	
  Ewing,	
  W.	
  R.,	
  and	
  Robl,	
  J.	
  A.	
  (2007)	
  Potent	
  and	
  
selective	
   biphenyl	
   azole	
   inhibitors	
   of	
   adipocyte	
   fatty	
   acid	
   binding	
   pirotein	
   (aFABP),	
  
Bioorganic	
  &	
  Medicinal	
  Chemistry	
  Letters	
  17,	
  3511-­‐3515.	
  
Yapici,	
   I.,	
   Lee,	
   K.	
   S.	
   S.,	
   Berbasova,	
   T.,	
   Nosrati,	
   M.,	
   Jia,	
   X.	
   F.,	
   Vasileiou,	
   C.,	
   Wang,	
   W.	
   J.,	
  
Santos,	
   E.	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2015)	
   "Turn-­‐On"	
   Protein	
   Fluorescence:	
   In	
  
Situ	
   Formation	
   of	
   Cyanine	
   Dyes,	
   Journal	
   of	
   the	
   American	
   Chemical	
   Society	
   137,	
   1073-­‐
1080.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  135,	
  16111-­‐16119.	
  
Abraham,	
  M.	
  H.,	
  Grellier,	
  P.	
  L.,	
  Abboud,	
  J.	
  L.	
  M.,	
  Doherty,	
  R.	
  M.,	
  and	
  Taft,	
  R.	
  W.	
  (1988)	
  
Solvent	
   Effects	
   in	
   Organic-­‐Chemistry	
   -­‐	
   Recent	
   Developments,	
   Canadian	
   Journal	
   of	
  
Chemistry-­‐Revue	
  Canadienne	
  De	
  Chimie	
  66,	
  2673-­‐2686.	
  
Ando,	
   Y.,	
   Homma,	
   Y.,	
   Hiruta,	
   Y.,	
   Citterio,	
   D.,	
   and	
   Suzuki,	
   K.	
   (2009)	
   Structural	
  
characteristics	
   and	
   optical	
   properties	
   of	
   a	
   series	
   of	
   solvatochromic	
   fluorescent	
   dyes	
  
displaying	
  long-­‐wavelength	
  emission,	
  Dyes	
  and	
  Pigments	
  83,	
  198-­‐206.	
  
Carlotti,	
   B.,	
   Flamini,	
   R.,	
   Kikas,	
   I.,	
   Mazzucato,	
   U.,	
   and	
   Spalletti,	
   A.	
   (2012)	
   Intramolecular	
  
charge	
   transfer,	
   solvatochromism	
   and	
   hyperpolarizability	
   of	
   compounds	
   bearing	
  
ethenylene	
  or	
  ethynylene	
  bridges,	
  Chemical	
  Physics	
  407,	
  9-­‐19.	
  

154	
  

113.	
  
	
  
114.	
  

	
  
109.	
  
	
  
110.	
  
	
  
111.	
  
	
  
112.	
  
113.	
  
	
  
114.	
  

	
  
109.	
  
	
  
110.	
  
	
  

Kamlet,	
  M.	
  J.,	
  and	
  Taft,	
  R.	
  W.	
  (1976)	
  Solvatochromic	
  Comparison	
  Method	
  1.	
  Beta-­‐Scale	
  
of	
  Solvent	
  Hydrogen-­‐Bond	
  Acceptor	
  (Hba)	
  Basicities,	
  Journal	
  of	
  the	
  American	
  Chemical	
  
Society	
  98,	
  377-­‐383.	
  
Yutani,	
   K.,	
   Ogasahara,	
   K.,	
   Tsujita,	
   T.,	
   and	
   Sugino,	
   Y.	
   (1987)	
   Dependence	
   of	
  
Conformational	
   Stability	
   on	
   Hydrophobicity	
   of	
   the	
   Amino-­‐Acid	
   Residue	
   in	
   a	
   Series	
   of	
  
Variant	
   Proteins	
   Substituted	
   at	
   a	
   Unique	
   Position	
   of	
   Tryptophan	
   Synthase	
   Alpha-­‐
Subunit,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  84,	
  4441-­‐4444.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  135,	
  16111-­‐16119.	
  
Abraham,	
  M.	
  H.,	
  Grellier,	
  P.	
  L.,	
  Abboud,	
  J.	
  L.	
  M.,	
  Doherty,	
  R.	
  M.,	
  and	
  Taft,	
  R.	
  W.	
  (1988)	
  
Solvent	
   Effects	
   in	
   Organic-­‐Chemistry	
   -­‐	
   Recent	
   Developments,	
   Canadian	
   Journal	
   of	
  
Chemistry-­‐Revue	
  Canadienne	
  De	
  Chimie	
  66,	
  2673-­‐2686.	
  
Ando,	
   Y.,	
   Homma,	
   Y.,	
   Hiruta,	
   Y.,	
   Citterio,	
   D.,	
   and	
   Suzuki,	
   K.	
   (2009)	
   Structural	
  
characteristics	
   and	
   optical	
   properties	
   of	
   a	
   series	
   of	
   solvatochromic	
   fluorescent	
   dyes	
  
displaying	
  long-­‐wavelength	
  emission,	
  Dyes	
  and	
  Pigments	
  83,	
  198-­‐206.	
  
Carlotti,	
   B.,	
   Flamini,	
   R.,	
   Kikas,	
   I.,	
   Mazzucato,	
   U.,	
   and	
   Spalletti,	
   A.	
   (2012)	
   Intramolecular	
  
charge	
   transfer,	
   solvatochromism	
   and	
   hyperpolarizability	
   of	
   compounds	
   bearing	
  
ethenylene	
  or	
  ethynylene	
  bridges,	
  Chemical	
  Physics	
  407,	
  9-­‐19.	
  
Kamlet,	
  M.	
  J.,	
  and	
  Taft,	
  R.	
  W.	
  (1976)	
  Solvatochromic	
  Comparison	
  Method	
  1.	
  Beta-­‐Scale	
  
of	
  Solvent	
  Hydrogen-­‐Bond	
  Acceptor	
  (Hba)	
  Basicities,	
  Journal	
  of	
  the	
  American	
  Chemical	
  
Society	
  98,	
  377-­‐383.	
  
Yutani,	
   K.,	
   Ogasahara,	
   K.,	
   Tsujita,	
   T.,	
   and	
   Sugino,	
   Y.	
   (1987)	
   Dependence	
   of	
  
Conformational	
   Stability	
   on	
   Hydrophobicity	
   of	
   the	
   Amino-­‐Acid	
   Residue	
   in	
   a	
   Series	
   of	
  
Variant	
   Proteins	
   Substituted	
   at	
   a	
   Unique	
   Position	
   of	
   Tryptophan	
   Synthase	
   Alpha-­‐
Subunit,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  84,	
  4441-­‐4444.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  135,	
  16111-­‐16119.	
  
Abraham,	
  M.	
  H.,	
  Grellier,	
  P.	
  L.,	
  Abboud,	
  J.	
  L.	
  M.,	
  Doherty,	
  R.	
  M.,	
  and	
  Taft,	
  R.	
  W.	
  (1988)	
  
Solvent	
   Effects	
   in	
   Organic-­‐Chemistry	
   -­‐	
   Recent	
   Developments,	
   Canadian	
   Journal	
   of	
  
Chemistry-­‐Revue	
  Canadienne	
  De	
  Chimie	
  66,	
  2673-­‐2686.	
  

155	
  

111.	
  
	
  
112.	
  
	
  
113.	
  
	
  
114.	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

Ando,	
   Y.,	
   Homma,	
   Y.,	
   Hiruta,	
   Y.,	
   Citterio,	
   D.,	
   and	
   Suzuki,	
   K.	
   (2009)	
   Structural	
  
characteristics	
   and	
   optical	
   properties	
   of	
   a	
   series	
   of	
   solvatochromic	
   fluorescent	
   dyes	
  
displaying	
  long-­‐wavelength	
  emission,	
  Dyes	
  and	
  Pigments	
  83,	
  198-­‐206.	
  
Carlotti,	
   B.,	
   Flamini,	
   R.,	
   Kikas,	
   I.,	
   Mazzucato,	
   U.,	
   and	
   Spalletti,	
   A.	
   (2012)	
   Intramolecular	
  
charge	
   transfer,	
   solvatochromism	
   and	
   hyperpolarizability	
   of	
   compounds	
   bearing	
  
ethenylene	
  or	
  ethynylene	
  bridges,	
  Chemical	
  Physics	
  407,	
  9-­‐19.	
  
Kamlet,	
  M.	
  J.,	
  and	
  Taft,	
  R.	
  W.	
  (1976)	
  Solvatochromic	
  Comparison	
  Method	
  1.	
  Beta-­‐Scale	
  
of	
  Solvent	
  Hydrogen-­‐Bond	
  Acceptor	
  (Hba)	
  Basicities,	
  Journal	
  of	
  the	
  American	
  Chemical	
  
Society	
  98,	
  377-­‐383.	
  
Yutani,	
   K.,	
   Ogasahara,	
   K.,	
   Tsujita,	
   T.,	
   and	
   Sugino,	
   Y.	
   (1987)	
   Dependence	
   of	
  
Conformational	
   Stability	
   on	
   Hydrophobicity	
   of	
   the	
   Amino-­‐Acid	
   Residue	
   in	
   a	
   Series	
   of	
  
Variant	
   Proteins	
   Substituted	
   at	
   a	
   Unique	
   Position	
   of	
   Tryptophan	
   Synthase	
   Alpha-­‐
Subunit,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  84,	
  4441-­‐4444.	
  

156	
  

	
  

Chapter	
  V:	
  Mutational	
  Studies	
  on	
  Starch	
  Branching	
  Enzyme	
  I	
  (RBEI)	
  to	
  understand	
  its	
  
Surface	
  Binding	
  Mechanism.	
  

V-­‐1	
  introduction	
  
Starch	
  is	
  the	
  world’s	
  second	
  most	
  abundant	
  polysaccharide.	
  More	
  than	
  1	
  billion	
  tons	
  of	
  
starch	
   is	
   consumed	
   annually,	
   as	
   it	
   provides	
   the	
   majority	
   of	
   caloric	
   intake	
   for	
   humans	
   and	
  
animals.	
   It	
   is	
   also	
   a	
   renewable	
   source	
   of	
   transportation	
   fuels.	
   It	
   is	
   therefore	
   critical	
   to	
  
understand	
   the	
   structure,	
   function	
   and	
   specificity	
   of	
   the	
   enzymes	
   responsible	
   for	
   starch	
  
biosynthesis	
  and	
  modification(118).	
  	
   Plants	
  use	
  exclusively	
  branched	
  linear	
  amylose	
  as	
  a	
  way	
  to	
  
store	
  excess	
  glucose.	
  The	
  stored	
  glucose	
  is	
  utilized	
  when	
  the	
  demand	
  for	
  energy	
  increases	
  or	
  
the	
   plant	
   is	
   unable	
   to	
   produce	
   enough	
   energy	
   to	
   maintain	
   e	
   functions.	
   The	
   polymer	
   is	
   stored	
  
either	
  in	
  special	
  organelles	
  called	
  amyloplasts,	
  found	
  in	
  tubers,	
  seeds	
  and	
  the	
  roots	
  of	
  plants,	
  or	
  
in	
  chlorplasts	
  in	
  the	
  leaf.	
  	
  
	
  	
  

In	
   vivo,	
   starch	
   synthase	
   is	
   responsible	
   synthesizing	
   the	
   linear	
   glucan	
   polymer	
   using	
   ADP-­‐

glucose	
  building	
  blocks.	
  The	
  final	
  step	
  in	
  the	
  starch	
  biosynthetic	
  pathway	
  	
  is	
  branching	
  enzyme	
  
catalyzes	
   the	
   branching	
   of	
   the	
   polysaccharide	
   by	
   cleaving	
   the	
   previously	
   formed	
   linear	
   α-­‐1,4	
  
glycosidic	
   linkage	
   by	
   SN2	
   attack	
   of	
   the	
   carboxylate	
   nucleophile	
   to	
   form	
   an	
   enzyme/substrate	
  
covalent	
   intermediat,	
   followed	
   by	
   a	
   second	
   SN2	
   reaction	
   by	
   the	
   acceptor	
   nucleophile	
   to	
  
reattaching	
  the	
  chain	
  via	
  an	
  α-­‐1,6	
  bond	
  that	
  serves	
  as	
  a	
  branching	
  point	
  on	
  the	
  newly	
  formed	
  
amylopectin(119).	
   These	
   polysaccharides	
   preserved	
   until	
   energy	
   stores	
   are	
   low,	
   at	
   which	
   time	
  
they	
   are	
   enzymatically	
   cleaved	
   from	
   the	
   main	
   chain	
   and	
   then	
   easily	
   metabolized	
   by	
   the	
  
organism.	
   	
   Fungi,	
   humans,	
   bacteria	
   and	
   other	
   animals	
   store	
   excess	
   glucose	
   as	
   glycogen	
   (a	
  
similarly	
  branched	
  polysaccharide)	
  in	
  an	
  analogous	
  fashion.	
  	
  	
  
157	
  

Branching	
   Enzymes	
   (BEs)	
   are	
   classified	
   as	
   belonging	
   to	
   glycoside	
   hydrolase	
   (GH)	
   family	
   13	
  
according	
   to	
   the	
   CAZy	
   classification	
   (http://www.cazy.org/)	
   based	
   on	
   similarity	
   in	
   the	
   amino	
  
acid	
   sequence,	
   typified	
   by	
   α-­‐amylase	
   and	
   related	
   enzymes,	
   such	
   as	
   isoamylase,	
   cyclodextrin	
  
glycosyltransferase	
   (CGTase)	
   and	
   pullulanase.	
   Members	
   of	
   the	
   family	
   consist	
   of	
   a	
   catalytic	
  
central	
  (β/ι)8	
  barrel	
  domain	
  that	
  contains	
  the	
  active	
  site	
  residues.	
  (120,	
  121)	
  
To	
   date,	
   starch	
   branching	
   enzymes	
   (SBEs)	
   have	
   been	
   characterized	
   from	
   multiple	
  
sources	
   including	
   maize,	
   (122-­‐131)	
   rice,	
   (132-­‐137)	
   wheat,	
   (138-­‐141)	
   barley,	
   (142-­‐144)	
   potato,	
   (145,	
   146)	
  
cassava,	
   (147)	
   pea	
   and	
   soybean.	
   (148-­‐150)	
   	
   In	
   rice	
   endosperm,	
   there	
   are	
   three	
   isoforms	
   of	
   the	
  
branching	
   enzymes,	
   BEI,	
   BEIIa,	
   and	
   BEIIb.	
   BEI	
   and	
   BEIIb	
   are	
   typically	
   specifically	
   expressed	
   in	
  
endosperm	
  while	
  BEIIa	
  is	
  ubiquitous	
  in	
  every	
  tissue	
  and	
  it	
  is	
  been	
  shown	
  that	
  BEIIa	
  is	
  required	
  
for	
   leaf	
   starch	
   synthesis	
  

(133,	
   137)

	
   suggesting	
   a	
   different	
   biochemical	
   role.	
   Studies	
   on	
   the	
  

properties	
  of	
  rice	
  BE	
  isozymes	
  (137)	
  supports	
  a	
  previous	
  observation	
  stating	
  that	
  the	
  branching	
  of	
  
the	
   amylopectin	
   amorphous	
   lamellae	
   is	
   catalyzed	
   by	
   BEI	
   while	
   the	
   branches	
   residing	
   on	
   the	
  
border	
   between	
   the	
   amorphous	
   and	
   crystalline	
   lamellae	
   are	
   almost	
   exclusively	
   catalyzed	
   by	
  
BEIIb.	
  (151)	
  Both	
  BEI	
  and	
  BEIIb	
  have	
  different	
  substrate	
  preferences.	
  
BEI	
  plays	
  a	
  role	
  in	
  transferring	
  less	
  branched,	
  longer	
  chains	
  with	
  a	
  degree	
  of	
  polymerization	
  (DP)	
  
>	
  37,but	
  most	
  preferably	
  those	
  with	
  degree	
  of	
  polymerization	
  DP 6-15	
  to	
  link	
  multiple	
  clusters	
  
of	
   amylopectin	
   5f,	
   while	
   BEIIb	
   specifically	
   transfers	
   short	
   chains	
   with	
   mostly	
   DP	
   of	
   6	
   and	
   7.	
  
(Figure	
   V-­‐1)	
   In	
   contrast,	
   the	
   BEIIa-­‐deficient	
   mutant	
   exhibited	
   no	
   significant	
   change	
   in	
   the	
  
amylopectin	
  chain.	
  (152)	
  The	
  physiological	
  role	
  of	
  BEIIa	
  might	
  be	
  to	
  support	
  at	
  least	
  partially	
  the	
  
function	
  of	
  BEI	
  and	
  BEIIb.(151),	
  5f	
  
	
  

The	
   structure	
   of	
   truncated	
   Oryza	
   Sativa	
   L	
   (Asian	
   rice)	
   branching	
   enzyme	
   (RBEI)	
   by	
  

158	
  

Noghuchi	
   et	
   al.	
   revealed	
   the	
   catalytic	
   residues	
   and	
   a	
   few	
   glucan	
   surface	
   binding	
   sites	
   (where	
  
the	
   protein	
   interacts	
   and	
   binds	
   with	
   amylose)	
   were	
   observed	
   to	
   play	
   an	
   important	
   role	
   on	
   how	
  
BEI	
  recognizes	
  polysaccharides	
  containing	
  α-­‐1,4	
  as	
  well	
  as	
  α-­‐1,6	
  linkages.	
  (2)	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
Figure	
  V-­‐1:	
  Schematic	
  view	
  of	
  the	
  BE	
  isozymes	
  and	
  SS	
  isozymes	
  in	
  the	
  amylopectin	
  cluster.	
  
	
  

V-­‐2	
  Structure	
  of	
  the	
  both	
  N	
  and	
  C	
  terminus	
  truncated	
  RBEI:	
  
The	
   crystal	
   structure	
   of	
   the	
   both	
   N	
   and	
   C	
   terminus	
   truncated	
   RBEI	
   first	
   has	
   been	
  
determined	
   by	
   Noghuchi	
   et	
   al.	
   in	
   2011.	
   Also,	
   Dr.	
   Remie	
   Fawaz	
   (a	
   former	
   Geiger	
   lab	
   member)	
  

159	
  

has	
   obtained	
   another	
   truncated	
   structure	
   of	
   this	
   enzyme	
   recently	
   with	
   dodecaose	
   (M12).	
  
(FigureV-­‐2)	
  	
  
BEI	
  is	
  820-­‐residues.	
  The	
  N-­‐terminal	
  leader	
  sequence	
  (65	
  residues),	
  which	
  is	
  responsible	
  
for	
   transport	
   into	
   the	
   amyloplast,	
   is	
   truncated	
   in	
   the	
   structure	
   and	
   leads	
   to	
   mature	
   BEI.	
   A	
  
mature	
   BEI	
   consists	
   of	
   755	
   residues	
   and	
   four	
   domains:	
   a	
   central	
   (β/ι)8	
   catalytic	
   domain	
   (the	
  
GH13	
   module,	
   residues	
   161–	
   587),	
   the	
   N-­‐terminal	
   carbohydrate-­‐binding	
   module	
   48	
   domain	
  
(CBM48;	
  residues	
  59–160),	
  the	
  N-­‐terminal	
  helices	
  (residues	
  1-­‐58)	
  and	
  an	
  α-­‐amylase	
  C-­‐	
  domain	
  
(residues	
   588–702)	
   (FigureI-­‐2).	
   Asp344	
   and	
   Glu399,	
   the	
   essential	
   catalytic	
   residues,	
   are	
   the	
  
nucleophile	
  (base)	
  and	
  proton	
  donor	
  respectively.	
  (2,	
  153)	
  	
  	
  
Crystal	
   structures	
   of	
   both	
   rice	
   mature	
   BEI	
   and	
   BEIΔC	
   (residues	
   702-­‐755	
   are	
   truncated)	
  
were	
  obtained	
  before.	
  (118)	
  In	
  our	
  group	
  	
  Dr.	
  Fawaz	
  was	
  able	
  to	
  obtain	
  well	
  diffracting	
  crystals	
  of	
  
malto-­‐dodecaose	
  (M12)	
  	
  bound	
  BEIΔC	
  at	
  2.35	
  Å.	
  
	
  
	
  
	
  

160	
  

Active	
  Site	
  

Binding	
  Site	
  I	
  

Binding	
  Site	
  II	
  
Figure	
   V-­‐2:	
   Overall	
   structure	
   of	
   rice	
   Branching	
   Enzyme	
   I	
   in	
   complex	
   with	
   maltododecaose	
  
(M12).	
   The	
   N-­‐terminal	
   domain	
   is	
   shown	
   in	
   green,	
   CBM48,	
   pink,	
   center	
   catalytic	
   domain,	
  
cyan,	
   and	
   the	
   C-­‐terminal	
   α-­‐amylase	
   C	
   domain	
   in	
   blue.	
   Carbohydrates	
   are	
   represented	
   in	
  
sticks	
   and	
   colored	
   by	
  atom	
   type,	
   carbons	
   in	
   yellow	
   and	
   oxygens	
   in	
   red.	
   One	
   oligosaccharide,	
  
M12,	
  binds	
  exclusively	
  into	
  the	
   catalytic	
  domain	
  and	
  hangs	
  over	
  the	
   catalytic	
  groove	
  without	
  
reaching	
  inside	
  (site	
  II),	
  while	
  the	
  five	
  glucose	
  units	
  visible	
  for	
  the	
  second	
  molecules	
  (site	
  I),	
  
between	
  three	
   domains:	
   the	
  N-­‐terminal,	
   the	
  carbohydrate	
   binding	
  module,	
   and	
  the	
   catalytic	
  
domain.	
  
	
  
	
  
	
  
	
  
	
  

161	
  

	
  

	
  

	
  

ACTIVE	
  CENTER	
  

	
  

	
  

Binding	
  site	
  I	
  

	
  
Binding	
  site	
  II	
  

	
  

Figure	
   V-­‐3:	
   Surface	
   depiction	
   of	
   RBEI	
   in	
   complex	
   with	
   M12.	
   At	
   the	
   center	
   of	
   this	
   groove,	
  
residues	
  involved	
   in	
  catalysis	
  are	
  shown	
   in	
  blue.	
  These	
   catalytic	
  residues,	
   Y235,	
  D270,	
   H275,	
  
R342,	
  D344,	
  E399,	
  H467	
  and	
  D468	
  according	
  to	
  RBEI	
  sequence	
  numbering,	
  (2)	
  were	
  predicted	
  
based	
  on	
  biochemical	
  and	
  structural	
  data	
  of	
  α-­‐amylase.	
  (3,	
  4)	
  	
  

V-­‐2-­‐1	
  Binding	
  Site	
  I:	
  Carbohydrate	
  Binding	
  Module	
  (CBM)	
  	
  
	
  

This	
   binding	
   site	
   has	
   been	
   observed	
   in	
   both	
   the	
   truncated	
   structures	
   of	
   RBEI,	
   with	
  

maltopentaose	
   (M5)	
   and	
   maltododecaose	
   (M12).	
   Based	
   on	
   the	
   crystal	
   structure,	
   this	
   site	
  
bridges	
   the	
   CBM48,	
   N	
   terminal	
   helical	
   and	
   catalytic	
   domains.	
   In	
   both	
   structures,	
   five	
   units	
   of	
  
glucose	
  are	
  visible	
  in	
  this	
  site	
  I.	
  The	
  saccharides	
  make	
  several	
  hydrogen-­‐bond	
  interactions	
  with	
  
the	
  residues	
  in	
  that	
  side	
  along	
  with	
  the	
  water	
  networks.	
  (Figure	
  V-­‐4).	
  	
  	
  
The	
  current	
  rice	
  Branching	
  Enzyme	
  I	
  structure	
  confirms	
  once	
  again	
  the	
  necessity	
  of	
  the	
  CBM48	
  
domain	
  for	
  carbohydrate	
  binding.	
  (TableV-­‐1).	
  

162	
  

Tr
p
72
His 294

Glu
320
Glu
Trp 320
72

Trp
319

Arg
323
Glu 45

Glu
295
3

His 44

Pro
74
4

	
  

	
  

Lys
97

2
Phe 100

5
Lys 99

1

	
  

Figure	
   V-­‐4:	
   Binding	
   site	
   I:	
   detailed	
   interactions	
   between	
   the	
   oligosaccharide	
   and	
   RBEI.	
   The	
  
protein	
  atoms	
  are	
  colored	
  by	
  type:	
  C	
  in	
  blue	
  marine,	
  O	
  in	
  red	
  and	
  N	
  in	
  dark	
  blue.	
  M12	
  atoms	
  are	
  
also	
   colored	
   by	
   type	
   with	
   C	
   in	
   yellow	
   and	
   O	
   in	
   red.	
   Glucose	
   units	
   are	
   numbered	
   in	
   red.	
  
Hydrogen	
   bonds	
   are	
   shown	
   in	
   dotted	
   black	
   lines.	
   Water	
   molecules	
   interacting	
   in	
   this	
   site	
   are	
  
represented	
  in	
  spheres	
  and	
  colored	
  in	
  cyan.	
  

V-­‐2-­‐2	
  New	
  Observed	
  Binding	
  Site	
  II	
  
	
  

The	
   structure	
   of	
   M12-­‐bound	
   RBEI	
   reveals	
   a	
   second	
   M12	
   molecule	
   bound.	
   The	
   second	
  

molecule	
   of	
   oligosaccharide	
   and	
   close	
   residues	
   around	
   it,	
   are	
   divided	
   into	
   two	
   binding	
   sites.	
  
Binding	
  site	
  II	
  are	
  starting	
  away	
  from	
  the	
  active	
  site.	
  (Figure	
  V-­‐2)	
  
Almost	
  all	
  glucose	
  units	
  in	
  this	
  carbohydrate	
  make	
  interactions	
  (specially	
  hydrogen	
  binding)	
  to	
  
residues	
  of	
  the	
  enzyme.	
  (Figure	
  V-­‐6	
  and	
  Table	
  V-­‐1)	
  
	
  

163	
  

	
  
Lys
484
Met
490

Thr
488

Met
486

Asp
483

5
Lys
475

6

10
8

12

Leu
562
His
561

Phe
479

7

9

11

	
  

Tyr
487

Ser
491
Leu
493

Pro
533

Tyr
564
Leu
556

	
  

3

4

Gln
553

Glu
534

Gly
473

2

	
  
1

Trp
535

Val
472

Ser
470
Asp
537

Phe
538

Arg
540

Figure	
  V-­‐5:	
  Binding	
  site	
  II,	
  detailed	
  interactions	
  between	
  the	
  oligosaccharide	
  and	
  RBEI.	
  Numbers	
  
for	
   glucose	
   units,	
   and	
   the	
   protein	
   atoms	
   are	
   colored.	
   Hydrogen	
   bonds	
   and	
   water	
   molecules	
   are	
  
also	
  represented.	
  Hydrogen	
  bonds	
  are	
  shown	
  in	
  dotted	
  black	
  lines.	
  

V-­‐3	
  Mutational	
  studies	
  on	
  the	
  observed	
  binding	
  sites	
  and	
  active	
  site	
  of	
  RBEI	
  
Several	
  essential	
  binding	
  sites	
  and	
  conserved	
  amino	
  acid	
  residues	
  are	
  now	
  revealed	
  in	
  
RBEI.	
  Therefore	
  in	
  order	
  to	
  understand	
  the	
  effect	
  of	
  each	
  residue	
  on	
  the	
  protein	
  activity	
  and	
  its	
  
involvement	
  in	
  determination	
  of	
  chain	
  length	
  transferred,	
  it	
  becomes	
  important	
  to	
  perform	
  a	
  
series	
   of	
   mutations	
   starting	
   with	
   point	
   mutations,	
   and	
   then	
   after	
   comparing	
   results	
   carrying	
  
double	
  mutants	
  of	
  RBEI.	
  	
  
	
  
	
  

164	
  

Table	
  V-­‐1:	
  Residues	
  involved	
  in	
  the	
  interactions	
  in	
  Binding	
  sites	
  I,	
  II	
  and	
  III	
  of	
  RBEI-­‐M12.	
  
	
  
Binding	
  site	
  

Residues	
  involved	
  in	
  binding	
  

	
  
	
  
I	
  

	
  
H44,	
  E45,	
  W72,	
  P74,	
  K97,	
  K99,	
  F100,	
  W319,	
  H294,	
  

	
  
	
  

E295,	
  E320,	
  R323	
  

	
  
	
  
II	
  

H561,	
  S491,	
  Y487,	
  T488,	
  M490,	
  Y564,	
  M486,	
  Q553,	
  
E534,	
   L493,	
   L562,	
   L556,	
   P533,	
   K484,	
   D483,	
   K475,	
  
G473,	
  S470,	
  D537,	
  F538,	
  R540,	
  F479,	
  W535,	
  V472	
  

	
  
	
  
	
  
	
  
	
  

V-­‐3-­‐1	
  Effect	
  of	
  Active	
  Site	
  Mutation:	
  
	
  	
  

As	
   previously	
   explained	
   in	
   section	
   V-­‐2,	
   Asp344	
   is	
   the	
   nucleophile	
   in	
   the	
   	
   catalytic	
  

process.	
   (Figure	
   V-­‐3,	
  Figure	
  V-­‐6)	
   to	
   confirm	
   this	
   the	
   D344A	
   mutant	
  was	
   made	
   to	
   study	
   its	
   effect	
  
on	
  the	
  enzyme	
  activity	
  and	
  function.	
  The	
  goal	
  is	
  to	
  confirm	
  its’	
  role	
  as	
  nucleophile,	
  and	
  to	
  use	
  
the	
   in-­‐active	
   enzyme	
   in	
   crystallization	
   experiments,	
   because	
   the	
   inactive	
   enzyme	
   may	
   bind	
   a	
  
glucan	
  in	
  the	
  active	
  site	
  without	
  degradation.	
  

165	
  

	
  
	
  
	
  

Active	
  site

	
  

Glu
399

Binding	
  site	
  I

Asp
344

	
  
	
  
	
  
	
  

	
  
Binding	
  site	
  II

	
  
Figure	
   V-­‐6:	
   Overall	
   structure	
   of	
   RBEI.	
   Active	
   site	
   residues	
   are	
   shown	
   in	
   blue	
   sticks.	
   Two	
  
oligosaccharide	
   molecules	
   are	
   in	
   yellow	
   sticks.	
   Oxygen	
   atoms	
   are	
   in	
   red.	
   So	
   far,	
   we	
   have	
   not	
  
identified	
  any	
  glucose	
  units	
  close	
  and	
  around	
  the	
  active	
  site	
  yet.	
  
	
  
The	
   iodine-­‐activity	
   assay	
   was	
   carried	
   out	
   on	
   purified	
   mutated	
   enzyme.	
   Please	
   refer	
   to	
  
section	
   V-­‐4-­‐1	
   and	
   V-­‐4-­‐2	
   for	
   full	
   details	
   on	
   the	
   purification	
   step	
   and	
   the	
   assay	
   protocol.	
   The	
  
absorbance	
  of	
  the	
  glucan–iodine	
  complex	
  was	
  almost	
  constant	
  during	
  this	
  time,	
  which	
  indicates	
  
the	
  deactivation	
  of	
  enzyme	
  due	
  to	
  the	
  mutation	
  in	
  the	
  active	
  site.	
  (Figure	
  V-­‐7)	
  Specific	
  acitivy	
  of	
  
D344A	
  was	
  0.02	
  μmol/min-­‐1mg-­‐1	
  (U/mg)	
  which	
  was	
  only	
  0.7%	
  active	
  compared	
  to	
  the	
  activity	
  of	
  
wild-­‐type	
   (WT)	
   truncated	
   RBEI.	
   Data	
   confirmed	
   that	
   we	
   knocked	
   out	
   the	
   enzyme	
   activity	
   by	
  
mutating	
  its	
  active	
  site	
  as	
  we	
  predicited.	
  	
  
Crystallization	
  attempts	
  with	
  this	
  mutant	
  have	
  so	
  far	
  not	
  yielded	
  well-­‐diffracting	
  crystals.	
  	
  

	
  

166	
  

Absorbance	
  at	
  660nm	
  (a.	
  u.)	
  

	
  
	
  
	
  
D344A	
  Acmvity	
  Assay	
  	
  
1.6	
  
	
  
	
  
1.4	
  
	
  
	
  
1.2	
  
	
  
1	
  
	
  
	
  
0.8	
  
D344A	
  	
  
	
  
y	
  =	
  -­‐0.0006x	
  +	
  1.4839	
  
	
  
0.6	
  
	
  
	
  
0.4	
  
	
  
	
  
0.2	
  
	
  
0	
  
	
  
0	
  
5	
  
10	
  
15	
  
20	
  
25	
  
30	
  
35	
  
	
  
Time	
  (min)	
  
	
  
Figure	
  V-­‐7:	
  Absorbance	
  versus	
  Time	
  for	
  the	
  D344A	
  mutation	
  of	
  truncated	
  RBEI.The	
  graph	
  is	
  the	
  
average	
  for	
  three	
  trials.	
  	
  Slope/Protein	
  indicates	
  the	
  activity	
  (U/mg)	
  of	
  the	
  enzyme.	
  
	
  

V-­‐3-­‐1	
  Effect	
  of	
  E534A	
  Mutation:	
  
In	
   binding	
   site	
   II	
   of	
   RBEI-­‐M12,	
   Glu	
   534	
   directly	
   interacts	
   with	
   Glucose7	
   via	
   a	
   hydrogen	
  
bond	
  to	
  O2	
  and	
  makes	
  a	
  water-­‐mediated	
  interaction	
  with	
  O3	
  (Figure	
  V-­‐5,	
  Figure	
  V-­‐8).	
  	
  

167	
  

10
7
	
  

	
  

9

11 	
  
12

6

	
  
8

Glu 534

Figure	
  V-­‐8:	
  Detailed	
  interactions	
  between	
  the	
  Glu7	
  in	
  binding	
  site	
  II	
  and	
  residue	
  E534	
  of	
  RBEI.	
  
All	
   interactions	
   between	
   them	
   are	
   in	
   between	
   2-­‐3.5	
   Å.	
   	
   Numbers	
   for	
   glucose	
   units,	
   and	
   the	
  
protein	
   atoms	
   are	
   colored.	
   Hydrogen	
   bonds	
   and	
   water	
   molecules	
   are	
   also	
   represented.	
  
Hydrogen	
  bonds	
  are	
  shown	
  in	
  dotted	
  black	
  lines.	
  	
  	
  
	
  
The	
  E534A	
  mutation	
  was	
  made	
  and	
  assayed	
  to	
  test	
  the	
  importance	
  of	
  this	
  interaction,	
  
which	
  is	
  located	
  far	
  from	
  the	
  active	
  site.	
  	
  However	
  the	
  iodine-­‐activity	
  assay	
  result	
  surprised	
  us	
  
and	
   E534A	
   mutant	
   was	
   83%	
   active	
   compare	
   to	
   the	
   WT	
   enzyme.	
   (Figure	
   V-­‐9)	
   Based	
   on	
   our	
  
result,	
   Glu	
   534	
   in	
   Site	
   II,	
   is	
   more	
   likely	
   to	
   involve	
   binding	
   enzyme	
   to	
   its	
   substrate	
   than	
   directing	
  
oligosaccharide	
   into	
   the	
   active	
   site	
   (because	
   mutation	
   of	
   this	
   residue,	
   cannot	
   deactivate	
   the	
  
enzyme).	
  	
  This	
  opens	
  up	
  the	
  idea	
  that	
  maybe	
  residues	
  around	
  that	
  site	
  are	
  most	
  likely	
  to	
  involve	
  
directing	
   transferring	
   chains	
   (donor)	
   and	
   acceptor	
   chains	
   into	
   the	
   active	
   site.	
   	
   To	
   confirm	
   our	
  
hypothesis	
  regarding	
  binding	
  site	
  II,	
  a	
  list	
  of	
  series	
  of	
  mutations	
  were	
  proposed	
  (Table	
  V-­‐2).	
  The	
  

168	
  

mutation’s	
   DNA	
   were	
   prepared	
   but	
   they	
   need	
   to	
   be	
   expressed	
   and	
   tested	
   for	
   their	
   activity	
   and	
  
crystallized	
  in	
  the	
  future.	
  
	
  

E534A	
  	
  Acmvity	
  Assay	
  
Absorbance	
  at	
  660nm	
  (a.	
  u.	
  )	
  

0.7	
  
0.6	
  
0.5	
  
0.4	
  
0.3	
  
0.2	
  

0.7	
  
0.6	
  
0.5	
  
0.4	
  
0.3	
  

E534A	
  

0.2	
  

y	
  =	
  -­‐0.0729x	
  +	
  0.628	
  
	
  

0.1	
  
0	
  

	
  

Absorbance	
  at	
  660nm	
  (a.	
  u.	
  )	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

0	
  

2	
  

4	
  

Time	
  (min)	
  

6	
  

0.1	
  
0	
  
0	
  

5	
  

10	
  

15	
  

20	
  

25	
  

30	
  

Time	
  (min)	
  
Figure	
  V-­‐9:	
  Absorbance	
   versus	
   Time	
   for	
  the	
  E534A	
  mutation	
  of	
  truncated	
  RBEI:	
  The	
   graph	
  is	
  
the	
   average	
   for	
   three	
   trials.	
   The	
   decrease	
   in	
   absorbance	
   is	
   linear	
   for	
   the	
   first	
   5	
   min	
   then	
  
reaches	
  a	
  plateau	
  due	
  to	
  the	
  high	
  activity	
  of	
  the	
  enzyme.	
  	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

169	
  

Table	
  V-­‐2:	
  	
  
More	
  suggested	
  mutations	
  to	
  study	
  for	
  residues	
  interacting	
  with	
  M12	
  in	
  binding	
  site	
  II	
  of	
  
RBEI.	
  

!
!
Residue'Number!

	
  
	
  
	
  

!
!
Residue!

!
Type%of%
interaction)
to#glucose(
units!

!
!
Mutation(!

561!

His!

Hydrophilic!

Ala!

562!

Leu!

Hydrophobic!

Phe!

564!

Tyr!

Hydrophilic!

Phe!

487!

Tyr!

Hydrophilic!

Ala!

	
  
	
  
	
  
V-­‐4	
  Experimental	
  
V-­‐4-­‐1	
  Material	
  and	
  Method	
  
Full-length RBEI (Oryza Sativa, Clone ID: 114619 from Japan) was cloned into a
modified pet28a vector (Novagen) by Dr. Fawaz. It contained N-terminus His6-tag and
SUMO site right before the N-terminus of RBEI (the first 65 residues were excluded).
There is also a tobacco etch virus (TEV) recognition site for cleavage in the C-terminus
of the RBEI sequence (between res 694-695).

To mutate the WT DNA, following

primers were ordered and site-directed mutagenesis was performed to mutate and
result in pure	
   DNA.	
   This	
   method	
   was	
   performed	
   by	
   pfuTutrbo	
   DNA	
   polymerase	
   (agilent)	
   and	
  
temperature	
  circle	
  in	
  PCR	
  instrument.	
  (Table	
  V-­‐3)	
  
	
  

170	
  

PCR	
  Protocol	
  for	
  RBEI	
  mutagenesis:	
  	
  
1x	
  	
  	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  95°C,	
  	
  3min	
  	
  

	
  
20x	
  
95°C	
  ,	
  	
  30	
  sec	
  
	
  
	
  
	
  	
  	
  	
  	
  
(Tm-­‐4)°C	
  ,	
  	
  3	
  min	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  
	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  	
  

72	
  °C	
  ,	
  	
  5min	
  

1x	
  	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  72°C,	
  10	
  min	
  	
  

1x	
  

	
  	
  	
  	
  	
  	
  	
  	
  	
  25°C,	
  	
  5	
  min	
  	
  

Tm	
   stands	
   for	
   the	
   melting	
   temperature	
   of	
   the	
   primer	
   for	
   each	
   mutation	
   and	
   it	
   is	
   mentioned	
   on	
  
each	
  ordered	
  primer	
  from	
  IDT.	
  	
  
	
  
PCR	
  Primers	
  for	
  D344A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  GGCTTCCGATTTGCTGGGGTTACGTCA	
  -­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGATGTAACCCCAGCAAATCGGAAGCC	
  -­‐3ʼ	
  
PCR	
  Primers	
  for	
  E534A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  TTTGGCCATCCAGCCTGGATTGACTTT-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  AAAGTCCAATCCAGGCTGGATGGCCAAA-­‐3ʼ	
  
PCR	
  Primers	
  for	
  D344A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  GGCTTCCGATTTGCTGGGGTTACGTCA	
  -­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGATGTAACCCCAGCAAATCGGAAGCC	
  -­‐3ʼ	
  
PCR	
  Primers	
  for	
  E534A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  TTTGGCCATCCAGCCTGGATTGACTTT-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  AAAGTCCAATCCAGGCTGGATGGCCAAA-­‐3ʼ	
  
PCR	
  Primers	
  for	
  H561A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  CGACACTGATGCCCTTCGATACA-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGTATCGAAGGGCATCAGTGTCG-­‐3ʼ	
  
171	
  

PCR	
  Primers	
  for	
  D344A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  GGCTTCCGATTTGCTGGGGTTACGTCA	
  -­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGATGTAACCCCAGCAAATCGGAAGCC	
  -­‐3ʼ	
  
PCR	
  Primers	
  for	
  E534A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  TTTGGCCATCCAGCCTGGATTGACTTT-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  AAAGTCCAATCCAGGCTGGATGGCCAAA-­‐3ʼ	
  
PCR	
  Primers	
  for	
  D344A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  GGCTTCCGATTTGCTGGGGTTACGTCA	
  -­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGATGTAACCCCAGCAAATCGGAAGCC	
  -­‐3ʼ	
  
PCR	
  Primers	
  for	
  E534A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  TTTGGCCATCCAGCCTGGATTGACTTT-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  AAAGTCCAATCCAGGCTGGATGGCCAAA-­‐3ʼ	
  
PCR	
  Primers	
  for	
  H561A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  CGACACTGATGCCCTTCGATACA-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGTATCGAAGGGCATCAGTGTCG-­‐3ʼ	
  
	
  
PCR	
  Primers	
  for	
  H561A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  CGACACTGATGCCCTTCGATACA-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGTATCGAAGGGCATCAGTGTCG-­‐3ʼ	
  
	
  
PCR	
  Primers	
  for	
  L562F	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  TACAATTCCTTGGCCATCAAGAGAA-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TTCTCTTGATGGCCAAGGAAATGTA-­‐3ʼ	
  
PCR	
  Primers	
  for	
  Y564F	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  GCATTCATATACTTGAATCGAAGGTGATCAG-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  CTGATCACCTTCGATTCAAGTATATGAATGC-­‐3ʼ	
  
PCR	
  Primers	
  for	
  H561A	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  CGACACTGATGCCCTTCGATACA-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TGTATCGAAGGGCATCAGTGTCG-­‐3ʼ	
  
	
  
PCR	
  Primers	
  for	
  L562F	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  TACAATTCCTTGGCCATCAAGAGAA-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  TTCTCTTGATGGCCAAGGAAATGTA-­‐3ʼ	
  
PCR	
  Primers	
  for	
  Y564F	
  mutant:	
  
Forward:	
  5ʼ-­‐	
  GCATTCATATACTTGAATCGAAGGTGATCAG-­‐3ʼ	
  
Reverse:	
  5ʼ-­‐	
  CTGATCACCTTCGATTCAAGTATATGAATGC-­‐3ʼ	
  

172	
  

Protein	
   over-­‐expression:	
   The	
   desired	
   DNA	
   of	
   RBEI	
   was	
   then	
   transformed	
   into	
   BL21(DE3)	
  
E.coli	
   cells.	
   The	
   cells	
   were	
   grown	
   in	
   LB	
   media,	
   after	
   the	
   OD600	
   reached	
   too	
   0.5-­‐0.6	
   induced	
   with	
  
0.5mM	
  IPTG	
  and	
  over	
  expressed	
  at	
  25°C	
  for	
  5	
  hours	
  before	
  centrifugation	
  and	
  freezing	
  at	
  -­‐20°C.	
  	
  	
  
The	
   frozen	
   cells	
   (obtained	
   from	
   6L)	
   were	
   re-­‐suspended	
   and	
   supplemented	
   with	
   one	
   table	
   of	
  
protease	
   inhibitor	
   tablet	
   in	
   lysis	
   buffer	
   (5mL/g	
   of	
   pellelts)	
   of	
   (50mM	
   Tris,	
   pH=8.0,	
   100mM	
   NaCl,	
  
1mM	
   BME,	
   10mM	
   imidazole).	
   The	
   re-­‐suspended	
   cells	
   were	
   lysed	
   by	
   sonication	
   (three	
   1	
   min	
  
with	
  1	
  min	
  relaxation	
  between	
  each)	
  and	
  centrifuged	
  for	
  20	
  min,	
  13000	
  rpm.	
  The	
  supernatant	
  
was	
  subjected	
  to	
  Ni-­‐NTA	
  affinity	
  resin	
  (from	
  Qiagen).	
  Resin	
  washed	
  with	
  the	
  same	
  buffer	
  plus	
  
30	
  mM	
  imidazole	
  until	
  it	
  reached	
  to	
  the	
  baseline.	
  Next,	
  Sumo	
  protease	
  added	
  on	
  the	
  column	
  
before	
   eluting	
   the	
   protein	
   and	
   it	
   was	
   collected	
   in	
   different	
   fractions.	
   30	
   ΟL	
   of	
   sample	
   is	
  
collected	
  for	
  running	
  polyacrylamide	
  gel	
  electrophoresis	
  (SDS-­‐PAGE)	
  at	
  each	
  step.	
  	
  
The	
  C-­‐terminus	
  cut	
  of	
  RBEI	
  was	
  carried	
  by	
  addition	
  of	
  the	
  TEV	
  protease	
  (Concentration	
  2	
  mg/mL)	
  
and	
  dialyzed	
  against	
  buffer	
  50mM	
  Tris-­‐HCL	
  pH	
  8.0,	
  0.5mM	
  EDTA,	
  1mM	
  DTT	
  for	
  overnight	
  at	
  4°C.	
  
Protein	
  was	
  concentrated	
  using	
  centrifugal	
  concentrator	
  with	
  10000	
  MW	
  cutoffs	
  and	
  purified	
  
on	
  a	
  Superdex	
  75	
  16/600	
  GE	
  column.	
  Fractions	
  with	
  right	
  size	
  of	
  truncated	
  RBEI	
  (80kDa)	
  were	
  
analyzed	
   by	
   12%	
   SDS	
   PAGE.	
   Pure	
   protein	
   then	
   was	
   concentrated	
   by	
   centrifugal	
   concentrator	
   to	
  
(3-­‐5	
  mg/ml),	
  as	
  determined	
  by	
  Bio-­‐Rad	
  protein	
  assay	
  at	
  600nm.	
  
V-­‐4-­‐2	
  Activity	
  Assay	
  
We	
   assayed	
   the	
   enzymatic	
   activity	
   for	
   the	
   truncated	
   RBEI	
   through	
   iodine-­‐staining	
   assay.	
  
Branching	
   enzymes	
   have	
   been	
   tested	
   for	
   their	
   activity	
   since	
   the	
   1970s	
   (131,	
   154,	
   155).	
   The	
   most	
  
commonly	
   used	
   assay,	
   was	
   developed	
   by	
   Boyer,	
   C.	
   and	
   Preiss,	
   J.	
   in	
   the	
   Biochemistry	
  
department	
   of	
   Michigan	
   State	
   University	
   (MSU).	
   This	
   assay	
   is	
   based	
   on	
   the	
   decrease	
   in	
  
173	
  

absorption	
   of	
   a	
   glucan-­‐iodine	
   complex.	
   (154).	
   The	
   absorbance	
   of	
   the	
   glucan–iodine	
   complex	
   is	
  
decreased	
   by	
   the	
   branching	
   of	
   the	
   substrate	
   (amylose	
   from	
   Sigma	
   Scientific)	
   with	
   branching	
  
enzyme.	
  There	
  are	
  3	
  trials	
  for	
  this	
  assay	
  simultaneously	
  (to	
  minimize	
  variables	
  in	
  substrates	
  or	
  
instrument	
   and	
   therefore	
   obtained	
   three	
   data	
   sets	
   per	
   protein)	
   all	
   contain	
   80𝜇L	
   of	
   amylose	
  
stock	
  solution	
  (0.5ml	
  10%NaOH,	
  50mg	
  amylose,	
  2ml	
  H2O),	
  6	
  𝜇L	
  HCL	
  and	
  820  𝜇L	
  H2O.	
  Solutions	
  
are	
  set	
  to	
  pH	
  8.0.	
  Tubes	
  stabilize	
  at	
  30°C	
  for	
  about	
  10	
  minutes.	
  The	
  iodine	
  reagent	
  was	
  made	
  
daily	
   from	
   2.9	
   mL	
   of	
   stock	
   solution	
   (0.26	
   g	
   of	
   I2	
   and	
   2.6g	
   of	
   KI	
   in	
   10mL	
   of	
   water).	
   Branching	
  
enzyme	
  I	
  (30  𝜇g)	
  is	
  added	
  to	
  each	
  tube	
  of	
  amylose	
  and	
  an	
  aliquot	
  of	
  50	
  ΟL	
  of	
  reaction	
  mixture	
  is	
  
withdrawn	
   to	
   add	
   to	
   the	
   iodine	
   solution	
   every	
   5	
   minutes	
   over	
   a	
   30	
   minute-­‐period.	
   The	
  
absorbance	
   was	
   determined	
   at	
   660	
   nm	
   for	
   the	
   reaction	
   mixtures	
   containing	
   amylose.	
   We	
  
monitor	
  the	
  change	
  in	
  absorbance	
  (y-­‐axis)	
  at	
  a	
  wavelength	
  of	
  660nm	
  in	
  5	
  minute	
  intervals	
  (X-­‐
axis)	
  over	
  a	
  30-­‐minute	
  total	
  reaction	
  time.	
  	
  The	
  greater	
  decrease	
  in	
  absorbance,	
  the	
  greater	
  the	
  
activity.	
  	
  

Activity (U / mg) =

Absorbance / min
slope
=
	
  
Protein Amount (mg)
0.03 mg

Percent Activity =

Protein Activity (U/mg)
× 100 	
  
WT Activity (U/mg)

In	
   the	
  case	
  of	
   E534A	
  Mutant,	
   the	
   plot	
  of	
   absorbance	
   versus	
   time	
   was	
   linear	
   for	
   the	
   first	
  
5	
   min,	
   but	
   then	
   reached	
   a	
   plateau.	
   Therefore,	
   we	
   used	
   the	
   linear	
   part	
   of	
   the	
   graph	
   to	
   calculate	
  
the	
  slope.	
  One	
  unit	
  of	
  activity	
  is	
  defined	
  as	
  a	
  decrease	
  in	
  absorbance	
  of	
  1.0	
  absorbance	
  unit	
  per	
  
min	
  at	
  660	
  nm	
  (154)	
  and	
  is	
  measured	
  in	
  U/mg	
  of	
  protein.	
  (154,	
  156)	
  
	
  

174	
  

V-­‐4-­‐3	
  Crystallization	
  of	
  RBEI	
  mutants	
  
	
  
Each	
  mutant	
  protein	
  was	
  concentrated	
  to	
  3-­‐5mg/mL.	
  Initial	
  screens	
  were	
  done	
  on	
  the	
  
mutant’s	
   proteins.	
   After	
   optimizing	
   different	
   conditions	
   for	
   crystals	
   to	
   grow,	
   the	
   best	
   condition	
  
was	
   28%	
   PEG	
   8000,	
   550mM	
   NaOAC,	
   0.1	
   M	
   (CH3)2AsO2Na	
   ·∙	
   3H2O	
   pH	
   6.9.	
   Few	
   small	
   crystals	
  
formed	
  after	
  a	
  month.	
  However	
  they	
  still	
  were	
  small	
  and	
  did	
  not	
  diffract.	
  	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  

	
  
	
  

175	
  

	
  
	
  
	
  
	
  
	
  
	
  
	
  
REFERENCES	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
	
  
176	
  

REFERENCES	
  
	
  
	
  
	
  
1.	
  
	
  
2.	
  

	
  
3.	
  
	
  
4.	
  

	
  
5.	
  
	
  
6.	
  

	
  
7.	
  
	
  
8.	
  
	
  
9.	
  

	
  
10.	
  

Harder,	
  M.	
  E.,	
  Deinzer,	
  M.	
  L.,	
  Leid,	
  M.	
  E.,	
  and	
  Schimerlik,	
  M.	
  I.	
  (2004)	
  Global	
  analysis	
  of	
  
three-­‐state	
  protein	
  unfolding	
  data,	
  Protein	
  Science	
  13,	
  2207-­‐2222.	
  
Noguchi,	
  J.,	
  Chaen,	
  K.,	
  Vu,	
  N.	
  T.,	
  Akasaka,	
  T.,	
  Shimada,	
  H.,	
  Nakashima,	
  T.,	
  Nishi,	
  A.,	
  Satoh,	
  
H.,	
   Omori,	
   T.,	
   Kakuta,	
   Y.,	
   and	
   Kimura,	
   M.	
   (2011)	
   Crystal	
   structure	
   of	
   the	
   branching	
  
enzyme	
  I	
  (BEI)	
  from	
  Oryza	
  sativa	
  L	
  with	
  implication	
  sfor	
  catalysis	
  and	
  susbtrate	
  binding,	
  
Glycobiology	
  21,	
  1108-­‐1116.	
  
Brzozowski,	
   A.	
   M.,	
   and	
   Davies,	
   G.	
   J.	
   (1997)	
   Structure	
   of	
   the	
   Aspergillus	
   oryzae	
   Alpha-­‐
Amylase	
   Complexed	
   with	
   the	
   Inhibitor	
   Acarbose	
   at	
   2.0	
   Å	
   Resolution,	
   Biochemistry	
   36,	
  
10837-­‐10845.	
  
Uitdehaag,	
  J.	
  C.	
  M.,	
  van	
  Alebeek,	
  G.	
  J.,	
  Van	
  Der	
  Veen,	
  B.,	
  Dijkhuizen,	
  L.,	
  and	
  Dijkstra,	
  B.	
  
W.	
  (2000)	
  Structures	
  of	
  Maltohexaose	
  and	
  Maltoheptaose	
  Bound	
  at	
  the	
  Donor	
  Sites	
  of	
  
Cyclodextrin	
  Glycosyltransferase	
  Give	
  Insight	
  into	
  the	
  Mechanisms	
  of	
  Transglycosylation	
  
Activity	
  and	
  Cyclodextrin	
  Size	
  Specificity,	
  Biochemistry	
  39,	
  7772-­‐7780.	
  
Ong,	
  D.	
  E.	
  (1984)	
  A	
  NOVEL	
  RETINOL-­‐BINDING	
  PROTEIN	
  FROM	
  RAT	
  -­‐	
  PURIFICATION	
  AND	
  
PARTIAL	
  CHARACTERIZATION,	
  Journal	
  of	
  Biological	
  Chemistry	
  259,	
  1476-­‐1482.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  crystallographica.	
  Section	
  
D,	
  Biological	
  crystallography	
  70,	
  3226-­‐3232.	
  
Veerkamp,	
   J.	
   H.,	
   Vankuppevelt,	
   T.,	
   Maatman,	
   R.,	
   and	
   Prinsen,	
   C.	
   F.	
   M.	
   (1993)	
  
STRUCTURAL	
   AND	
   FUNCTIONAL-­‐ASPECTS	
   OF	
   CYTOSOLIC	
   FATTY-­‐ACID-­‐BINDING	
  
PROTEINS,	
  Prostaglandins	
  Leukotrienes	
  and	
  Essential	
  Fatty	
  Acids	
  49,	
  887-­‐906.	
  
Banaszak,	
   L.,	
   Winter,	
   N.,	
   Xu,	
   Z.	
   H.,	
   Bernlohr,	
   D.	
   A.,	
   Cowan,	
   S.,	
   and	
   Jones,	
   T.	
   A.	
   (1994)	
  
LIPID-­‐BINDING	
   PROTEINS	
   -­‐	
   A	
   FAMILY	
   OF	
   FATTY-­‐ACID	
   AND	
   RETINOID	
   TRANSPORT	
  
PROTEINS,	
  Advances	
  in	
  Protein	
  Chemistry,	
  Vol	
  45	
  45,	
  89-­‐151.	
  
Vaezeslami,	
   S.,	
   Mathes,	
   E.,	
   Vasilelou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2006)	
   The	
  
structure	
   of	
   apo-­‐wild-­‐type	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   at	
   1.4	
   A	
   and	
   its	
  
relationship	
   to	
   ligand	
   binding	
   and	
   nuclear	
   translocation,	
   Journal	
   of	
   Molecular	
   Biology	
  
363,	
  687-­‐701.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins-­‐-­‐insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  lipid	
  research	
  43,	
  328-­‐349.	
  
177	
  

11.	
  
	
  
12.	
  
	
  
13.	
  
	
  
14.	
  

	
  
15.	
  

	
  
16.	
  

	
  
17.	
  

	
  
18.	
  
	
  
19.	
  
	
  
20.	
  

Lucke,	
   C.,	
   Huang,	
   S.,	
   Rademacher,	
   M.,	
   and	
   Ruterjans,	
   H.	
   (2002)	
   New	
   insights	
   into	
  
intracellular	
   lipid	
   binding	
   proteins:	
   The	
   role	
   of	
   buried	
   water,	
   Protein	
   science	
   :	
   a	
  
publication	
  of	
  the	
  Protein	
  Society	
  11,	
  2382-­‐2392.	
  
Schaap,	
   F.	
   G.,	
   van	
   der	
   Vusse,	
   G.	
   J.,	
   and	
   Glatz,	
   J.	
   F.	
   (2002)	
   Evolution	
   of	
   the	
   family	
   of	
  
intracellular	
   lipid	
   binding	
   proteins	
   in	
   vertebrates,	
   Molecular	
   and	
   cellular	
   biochemistry	
  
239,	
  69-­‐77.	
  
Smathers,	
   R.	
   L.,	
   and	
   Petersen,	
   D.	
   R.	
   (2011)	
   The	
   human	
   fatty	
   acid-­‐binding	
   protein	
   family:	
  
evolutionary	
  divergences	
  and	
  functions,	
  Human	
  genomics	
  5,	
  170-­‐191.	
  
Franzoni,	
   L.,	
   Lucke,	
   C.,	
   Perez,	
   C.,	
   Cavazzini,	
   D.,	
   Rademacher,	
   M.,	
   Ludwig,	
   C.,	
   Spisni,	
   A.,	
  
Rossi,	
  G.	
  L.,	
  and	
  Ruterjans,	
  H.	
  (2002)	
  Structure	
  and	
  backbone	
  dynamics	
  of	
  apo-­‐	
  and	
  holo-­‐
cellular	
   retinol-­‐binding	
   protein	
   in	
   solution,	
   Journal	
   of	
   Biological	
   Chemistry	
   277,	
   21983-­‐
21997.	
  
Crow,	
   J.	
   A.,	
   and	
   Ong,	
   D.	
   E.	
   (1985)	
   CELL-­‐SPECIFIC	
   IMMUNOHISTOCHEMICAL	
  
LOCALIZATION	
   OF	
   A	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   (TYPE-­‐2)	
   IN	
   THE	
   SMALL-­‐
INTESTINE	
  OF	
  RAT,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  
of	
  America	
  82,	
  4707-­‐4711.	
  
Vogel,	
  S.,	
  Mendelsohn,	
  C.	
  L.,	
  Mertz,	
  J.	
  R.,	
  Piantedosi,	
  R.,	
  Waldburger,	
  C.,	
  Gottesman,	
  M.	
  
E.,	
  and	
  Blaner,	
  W.	
  S.	
  (2001)	
  Characterization	
  of	
  a	
  new	
  member	
  of	
  the	
  fatty	
  acid-­‐binding	
  
protein	
   family	
   that	
   binds	
   all-­‐trans-­‐retinol,	
   Journal	
   of	
   Biological	
   Chemistry	
   276,	
   1353-­‐
1360.	
  
Folli,	
   C.,	
   Calderone,	
   V.,	
   Ottonello,	
   S.,	
   Bolchi,	
   A.,	
   Zanotti,	
   G.,	
   Stoppini,	
   M.,	
   and	
   Berni,	
   R.	
  
(2001)	
   Identification,	
   retinoid	
   binding,	
   and	
   x-­‐ray	
   analysis	
   of	
   a	
   human	
   retinol-­‐binding	
  
protein,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  98,	
  3710-­‐3715.	
  
Li,	
  E.,	
  Demmer,	
  L.	
  A.,	
  Sweetser,	
  D.	
  A.,	
  Ong,	
  D.	
  E.,	
  and	
  Gordon,	
  J.	
  I.	
  (1986)	
  ISOLATION	
  OF	
  A	
  
CLONED	
   CDNA-­‐ENCODING	
   AN	
   ABUNDANT	
   INTESTINAL	
   RETINOL	
   BINDING-­‐PROTEIN,	
  
Clinical	
  Research	
  34,	
  A685-­‐A685.	
  
Li,	
   E.,	
   and	
   Norris,	
   A.	
   W.	
   (1996)	
   Structure/function	
   of	
   cytoplasmic	
   vitamin	
   A-­‐binding	
  
proteins,	
  Annual	
  Review	
  of	
  Nutrition	
  16,	
  205-­‐234.	
  
Li,	
   E.,	
   Qian,	
   S.	
   J.,	
   Winter,	
   N.	
   S.,	
   Davignon,	
   A.,	
   Levin,	
   M.	
   S.,	
   and	
   Gordon,	
   J.	
   I.	
   (1991)	
  
FLUORINE	
   NUCLEAR-­‐MAGNETIC-­‐RESONANCE	
   ANALYSIS	
   OF	
   THE	
   LIGAND-­‐BINDING	
  
PROPERTIES	
   OF	
   2	
   HOMOLOGOUS	
   RAT	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEINS	
  
EXPRESSED	
  IN	
  ESCHERICHIA-­‐COLI,	
  Journal	
  of	
  Biological	
  Chemistry	
  266,	
  3622-­‐3629.	
  

178	
  

21.	
  
	
  
22.	
  
	
  
23.	
  
	
  
24.	
  
	
  
25.	
  
	
  
26.	
  

	
  
27.	
  
	
  
28.	
  
29.	
  
	
  
30.	
  

	
  
31.	
  

Calderone,	
   V.,	
   Folli,	
   C.,	
   Marchesani,	
   A.,	
   Berni,	
   R.,	
   and	
   Zanotti,	
   G.	
   (2002)	
   Identification	
  
and	
   structural	
   analysis	
   of	
   a	
   zebrafish	
   apo	
   and	
   holo	
   cellular	
   retinol-­‐binding	
   protein,	
  
Journal	
  of	
  Molecular	
  Biology	
  321,	
  527-­‐535.	
  
Tarter,	
  M.,	
  Capaldi,	
  S.,	
  Carrizo,	
  M.	
  E.,	
  Ambrosi,	
  E.,	
  Perduca,	
  M.,	
  and	
  Monaco,	
  H.	
  L.	
  (2008)	
  
Crystal	
   structure	
   of	
   human	
   cellular	
   retinol-­‐binding	
   protein	
   11	
   to	
   1.2	
   angstrom	
  
resolution,	
  Proteins-­‐Structure	
  Function	
  and	
  Bioinformatics	
  70,	
  1626-­‐1630.	
  
Winter,	
   N.	
   S.,	
   Bratt,	
   J.	
   M.,	
   and	
   Banaszak,	
   L.	
   J.	
   (1993)	
   CRYSTAL-­‐STRUCTURES	
   OF	
   HOLO	
  
AND	
   APO-­‐CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN-­‐II,	
   Journal	
   of	
   Molecular	
   Biology	
   230,	
  
1247-­‐1259.	
  
Kane,	
  M.	
  A.,	
  Bright,	
  F.	
  V.,	
  and	
  Napoli,	
  J.	
  L.	
  (2011)	
  Binding	
  affinities	
  of	
  CRBPI	
  and	
  CRBPII	
  
for	
  9-­‐cis-­‐retinoids,	
  Biochimica	
  Et	
  Biophysica	
  Acta-­‐General	
  Subjects	
  1810,	
  514-­‐518.	
  
Malpeli,	
   G.,	
   Stoppini,	
   M.,	
   Zapponi,	
   M.	
   C.,	
   Folli,	
   C.,	
   and	
   Berni,	
   R.	
   (1995)	
   INTERACTIONS	
  
WITH	
   RETINOL	
   AND	
   RETINOIDS	
   OF	
   BOVINE	
   CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN,	
  
European	
  Journal	
  of	
  Biochemistry	
  229,	
  486-­‐493.	
  
McBee,	
   J.	
   K.,	
   Kuksa,	
   V.,	
   Alvarez,	
   R.,	
   de	
   Lera,	
   A.	
   R.,	
   Prezhdo,	
   O.,	
   Haeseleer,	
   F.,	
   Sokal,	
   I.,	
  
and	
   Palczewski,	
   K.	
   (2000)	
   Isomerization	
   of	
   all-­‐trans-­‐retinol	
   to	
   cis-­‐retinols	
   in	
   bovine	
  
retinal	
   pigment	
   epithelial	
   cells:	
   Dependence	
   on	
   the	
   specificity	
   of	
   retinoid-­‐binding	
  
proteins,	
  Biochemistry	
  39,	
  11370-­‐11380.	
  
Borshchevskiy,	
  V.	
  I.,	
  Round,	
  E.	
  S.,	
  Popov,	
  A.	
  N.,	
  Bueldt,	
  G.,	
  and	
  Gordeliy,	
  V.	
  I.	
  (2011)	
  X-­‐
ray-­‐Radiation-­‐Induced	
   Changes	
   in	
   Bacteriorhodopsin	
   Structure,	
   Journal	
   of	
   Molecular	
  
Biology	
  409,	
  813-­‐825.	
  
Matsui,	
  Y.,	
  Sakai,	
  K.,	
  Murakami,	
  M.,	
  Shiro,	
  Y.,	
  Adachi,	
  S.,	
  Okumura,	
  H.,	
  and	
  Kouyama,	
  T.	
  
(2002)	
  Specific	
  damage	
  induced	
  by	
  X-­‐ray	
  radiation	
  and	
  structural	
  changes	
  in	
  the	
  primary	
  
photoreaction	
  of	
  bacteriorhodopsin,	
  Journal	
  of	
  Molecular	
  Biology	
  324,	
  469-­‐481.	
  
Schreckenbach,	
   T.,	
   Walckhoff,	
   B.,	
   and	
   Oesterhelt,	
   D.	
   (1977)	
   STUDIES	
   ON	
   RETINAL-­‐
PROTEIN	
  INTERACTION	
  IN	
  BACTERIORHODOPSIN,	
  European	
  Journal	
  of	
  Biochemistry	
  76,	
  
499-­‐511.	
  
Cowan,	
   S.	
   W.,	
   Newcomer,	
   M.	
   E.,	
   and	
   Jones,	
   T.	
   A.	
   (1993)	
   CRYSTALLOGRAPHIC	
   STUDIES	
  
ON	
   A	
   FAMILY	
   OF	
   CELLULAR	
   LIPOPHILIC	
   TRANSPORT	
   PROTEINS	
   -­‐	
   REFINEMENT	
   OF	
   P2	
  
MYELIN	
   PROTEIN	
   AND	
   THE	
   STRUCTURE	
   DETERMINATION	
   AND	
   REFINEMENT	
   OF	
  
CELLULAR	
   RETINOL-­‐BINDING	
   PROTEIN	
   IN	
   COMPLEX	
   WITH	
   ALL-­‐TRANS-­‐RETINOL,	
   Journal	
  
of	
  Molecular	
  Biology	
  230,	
  1225-­‐1246.	
  
Vasileiou,	
   C.,	
   Vaezeslami,	
   S.,	
   Crist,	
   R.	
   M.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2007)	
   Protein	
   design:	
   reengineering	
   cellular	
   retinoic	
   acid	
   binding	
   protein	
   II	
   into	
   a	
  
rhodopsin	
  protein	
  mimic,	
  J	
  Am	
  Chem	
  Soc	
  129,	
  6140-­‐6148.	
  
179	
  

32.	
  
	
  
33.	
  
	
  
34.	
  
	
  
35.	
  

	
  
36.	
  
	
  
37.	
  
	
  
38.	
  
	
  
39.	
  
	
  
40.	
  
	
  
41.	
  
	
  
42.	
  
	
  
43.	
  
	
  
44.	
  
	
  

Otwinowski,	
   Z.,	
   and	
   Minor,	
   W.	
   (1997)	
   Processing	
   of	
   X-­‐ray	
   diffraction	
   data	
   collected	
   in	
  
oscillation	
  mode,	
  Macromolecular	
  Crystallography,	
  Pt	
  A	
  276,	
  307-­‐326.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (1997)	
   MOLREP:	
   an	
   automated	
   program	
   for	
   molecular	
  
replacement,	
  Journal	
  of	
  Applied	
  Crystallography	
  30,	
  1022-­‐1025.	
  
Vagin,	
   A.,	
   and	
   Teplyakov,	
   A.	
   (2010)	
   Molecular	
   replacement	
   with	
   MOLREP,	
   Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  22-­‐25.	
  
Murshudov,	
   G.	
   N.,	
   Skubak,	
   P.,	
   Lebedev,	
   A.	
   A.,	
   Pannu,	
   N.	
   S.,	
   Steiner,	
   R.	
   A.,	
   Nicholls,	
   R.	
   A.,	
  
Winn,	
   M.	
   D.,	
   Long,	
   F.,	
   and	
   Vagin,	
   A.	
   A.	
   (2011)	
   REFMAC5	
   for	
   the	
   refinement	
   of	
  
macromolecular	
   crystal	
   structures,	
   Acta	
   Crystallographica	
   Section	
   D-­‐Biological	
  
Crystallography	
  67,	
  355-­‐367.	
  
Emsley,	
   P.,	
   Lohkamp,	
   B.,	
   Scott,	
   W.	
   G.,	
   and	
   Cowtan,	
   K.	
   (2010)	
   Features	
   and	
   development	
  
of	
  Coot,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  486-­‐501.	
  
Bennett,	
   M.	
   J.,	
   Schlunegger,	
   M.	
   P.,	
   and	
   Eisenberg,	
   D.	
   (1995)	
   3D	
   domain	
   swapping	
   -­‐	
   A	
  
mechanism	
  for	
  oligomer	
  assembly,	
  Protein	
  Science	
  4,	
  2455-­‐2468.	
  
Crestfield,	
   A.	
   M.,	
   Stein,	
   W.	
   H.,	
   and	
   Moore,	
   S.	
   (1962)	
   On	
   the	
   aggregation	
   of	
   bovine	
  
pancreatic	
  ribonuclease,	
  Archives	
  of	
  biochemistry	
  and	
  biophysics	
  Suppl	
  1,	
  217-­‐222.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   As	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  Science	
  11,	
  1285-­‐1299.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   Structures	
   of	
   the	
   two	
   3D	
  
domain-­‐swapped	
  RNase	
  A	
  trimers,	
  Protein	
  Science	
  11,	
  371-­‐380.	
  
Liu,	
   Y.	
   S.,	
   Gotte,	
   G.,	
   Libonati,	
   M.,	
   and	
   Eisenberg,	
   D.	
   S.	
   (2001)	
   Crystal	
   structure	
   of	
   an	
  
Rnase	
  A	
  dimer	
  displaying	
  a	
  new	
  type	
  of	
  3D	
  domain	
  swapping,	
  Nucleic	
  Acid	
  Database.	
  
Liu,	
  Y.	
  S.,	
  Hart,	
  P.	
  J.,	
  Schlunegger,	
  M.	
  P.,	
  and	
  Eisenberg,	
  D.	
  (1998)	
  The	
  crystal	
  structure	
  of	
  
a	
   3D	
   domain-­‐swapped	
   dimer	
   of	
   RNase	
   A	
   at	
   a	
   2.1-­‐angstrom	
   resolution,	
   Proceedings	
   of	
  
the	
  National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  95,	
  3437-­‐3442.	
  
Liu,	
   Y.,	
   and	
   Eisenberg,	
   D.	
   (2002)	
   3D	
   domain	
   swapping:	
   as	
   domains	
   continue	
   to	
   swap,	
  
Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  11,	
  1285-­‐1299.	
  
Koharudin,	
   L.	
   M.,	
   Liu,	
   L.,	
   and	
   Gronenborn,	
   A.	
   M.	
   (2013)	
   Different	
   3D	
   domain-­‐swapped	
  
oligomeric	
   cyanovirin-­‐N	
   structures	
   suggest	
   trapped	
   folding	
   intermediates,	
   Proc	
   Natl	
  
Acad	
  Sci	
  U	
  S	
  A	
  110,	
  7702-­‐7707.	
  

180	
  

45.	
  
	
  
46.	
  
	
  
47.	
  
	
  
48.	
  
	
  
49.	
  

	
  
50.	
  
	
  
51.	
  
	
  
52.	
  
	
  
53.	
  
	
  
54.	
  
	
  
55.	
  
	
  

Assar,	
  Z.,	
  Nossoni,	
  Z.,	
  Wang,	
  W.,	
  Santos,	
  E.	
  M.,	
  Kramer,	
  K.,	
  McCornack,	
  C.,	
  Vasileiou,	
  C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   Domain-­‐Swapped	
   Dimers	
   of	
   Intracellular	
   Lipid-­‐
Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure	
  24,	
  1590-­‐1598.	
  
Bax,	
  B.,	
  Lapatto,	
  R.,	
  Nalini,	
  V.,	
  Driessen,	
  H.,	
  Lindley,	
  P.	
  F.,	
  Mahadevan,	
  D.,	
  Blundell,	
  T.	
  L.,	
  
and	
   Slingsby,	
   C.	
   (1990)	
   X-­‐ray	
   analysis	
   of	
   beta-­‐B2-­‐crystallin	
   and	
   evolution	
   of	
   oligomeric	
  
lens	
  proteins,	
  Nature	
  347,	
  776-­‐780.	
  
Hakansson,	
  M.,	
  Svensson,	
  A.,	
  Fast,	
  J.,	
  and	
  Linse,	
  S.	
  (2001)	
  An	
  extended	
  hydrophobic	
  core	
  
induces	
  EF-­‐hand	
  swapping,	
  Protein	
  Science	
  10,	
  927-­‐933.	
  
Rousseau,	
   F.,	
   Schymkowitz,	
   J.,	
   and	
   Itzhaki,	
   L.	
   S.	
   (2012)	
   Implications	
   of	
   3D	
   domain	
  
swapping	
   for	
   protein	
   folding,	
   misfolding	
   and	
   function,	
   In	
   Protein	
   Dimerization	
   and	
  
Oligomerization	
  in	
  Biology	
  (Mattews,	
  J.	
  M.,	
  Ed.),	
  pp	
  137-­‐152.	
  
Yang,	
   F.,	
   Bewley,	
   C.	
   A.,	
   Louis,	
   J.	
   M.,	
   Gustafson,	
   K.	
   R.,	
   Boyd,	
   M.	
   R.,	
   Gronenborn,	
   A.	
   M.,	
  
Clore,	
  G.	
  M.,	
  and	
  Wlodawer,	
  A.	
  (1999)	
  Crystal	
  structure	
  of	
  cyanovirin-­‐N,	
  a	
  potent	
  HIV-­‐
inactivating	
  protein,	
  shows	
  unexpected	
  domain	
  swapping,	
  Journal	
  of	
  Molecular	
  Biology	
  
288,	
  403-­‐412.	
  
Szilagyi,	
   A.,	
   Zhang,	
   Y.,	
   and	
   Zavodszky,	
   P.	
   (2012)	
   Intra-­‐Chain	
   3D	
   Segment	
   Swapping	
  
Spawns	
  the	
  Evolution	
  of	
  New	
  Multidomain	
  Protein	
  Architectures,	
  Journal	
  of	
  Molecular	
  
Biology	
  415,	
  221-­‐235.	
  
Stroud,	
   J.	
   C.,	
   Liu,	
   C.,	
   Teng,	
   P.	
   K.,	
   and	
   Eisenberg,	
   D.	
   (2012)	
   Toxic	
   fibrillar	
   oligomers	
   of	
  
amyloid-­‐beta	
   have	
   cross-­‐beta	
   structure,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  109,	
  7717-­‐7722.	
  
Cleary,	
   J.	
   P.,	
   Walsh,	
   D.	
   M.,	
   Hofmeister,	
   J.	
   J.,	
   Shankar,	
   G.	
   M.,	
   Kuskowski,	
   M.	
   A.,	
   Selkoe,	
   D.	
  
J.,	
   and	
   Ashe,	
   K.	
   H.	
   (2005)	
   Natural	
   oligomers	
   of	
   the	
   amyloid-­‐protein	
   specifically	
   disrupt	
  
cognitive	
  function,	
  Nature	
  Neuroscience	
  8,	
  79-­‐84.	
  
Fandrich,	
   M.,	
   Fletcher,	
   M.	
   A.,	
   and	
   Dobson,	
   C.	
   M.	
   (2001)	
   Amyloid	
   fibrils	
   from	
   muscle	
  
myoglobin	
   -­‐	
   Even	
   an	
   ordinary	
   globular	
   protein	
   can	
   assume	
   a	
   rogue	
   guise	
   if	
   conditions	
  
are	
  right,	
  Nature	
  410,	
  165-­‐166.	
  
Gronenborn,	
   A.	
   M.	
   (2009)	
   Protein	
   acrobatics	
   in	
   pairs	
   -­‐	
   dimerization	
   via	
   domain	
  
swapping,	
  Current	
  Opinion	
  in	
  Structural	
  Biology	
  19,	
  39-­‐49.	
  
Schlunegger,	
  M.	
  P.,	
  Bennett,	
  M.	
  J.,	
  and	
  Eisenberg,	
  D.	
  (1997)	
  Oligomer	
  formation	
  by	
  3D	
  
domain	
   swapping:	
   A	
   model	
   for	
   protein	
   assembly	
   and	
   misassembly,	
   In	
   Advances	
   in	
  
Protein	
  Chemistry,	
  Vol	
  50:	
  Protein	
  Misassembly	
  (Wetzel,	
  R.,	
  Ed.),	
  pp	
  61-­‐122.	
  

181	
  

56.	
  
	
  
57.	
  
	
  
58.	
  
	
  
59.	
  
	
  
60.	
  
	
  
61.	
  
	
  
62.	
  
	
  
63.	
  
	
  
64.	
  
	
  
65.	
  
	
  
66.	
  
	
  
67.	
  
	
  
68.	
  

Selkoe,	
  D.	
  J.	
  (2004)	
  Folding	
  proteins	
  in	
  fatal	
  ways	
  (vol	
  426,	
  pg	
  900,	
  2003),	
  Nature	
  428,	
  
445-­‐445.	
  
Shtilerman,	
  M.	
  D.,	
  Ding,	
  T.	
  T.,	
  and	
  Lansbury,	
  P.	
  T.	
  (2002)	
  Molecular	
  crowding	
  accelerates	
  
fibrillization	
   of	
   alpha-­‐synuclein:	
   Could	
   an	
   increase	
   in	
   the	
   cytoplasmic	
   protein	
  
concentration	
  induce	
  Parkinson's	
  disease?,	
  Biochemistry	
  41,	
  3855-­‐3860.	
  
Sunde,	
  M.,	
  and	
  Blake,	
  C.	
  (1997)	
  The	
  structure	
  of	
  amyloid	
  fibrils	
  by	
  electron	
  microscopy	
  
and	
   X-­‐ray	
   diffraction,	
   In	
   Advances	
   in	
   Protein	
   Chemistry,	
   Vol	
   50:	
   Protein	
   Misassembly	
  
(Wetzel,	
  R.,	
  Ed.),	
  pp	
  123-­‐159.	
  
Bennett,	
  M.	
  J.,	
  Choe,	
  S.,	
  and	
  Eisenberg,	
  D.	
  (1994)	
  Refined	
  structure	
  of	
  dimeric	
  diphtheria	
  
toxin	
  at	
  2.0	
  A	
  resolution,	
  Protein	
  science	
  :	
  a	
  publication	
  of	
  the	
  Protein	
  Society	
  3,	
  1444-­‐
1463.	
  
O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
  environment,	
  Curr	
  Opin	
  Struct	
  Biol	
  11,	
  114-­‐119.	
  
Bennett,	
   M.	
   J.,	
   Choe,	
   S.,	
   and	
   Eisenberg,	
   D.	
   (1994)	
   Rlefined	
   strucutre	
   of	
   dimeric	
  
diphtheria-­‐toxin	
  at	
  2.0-­‐angstrom	
  resolution,	
  Protein	
  Science	
  3,	
  1444-­‐1463.	
  
Carroll,	
   S.	
   F.,	
   Barbieri,	
   J.	
   T.,	
   and	
   Collier,	
   R.	
   J.	
   (1986)	
   Dimeric	
   form	
   of	
   diphtheria-­‐toxin-­‐
purification	
  and	
  characterization,	
  Biochemistry	
  25,	
  2425-­‐2430.	
  
Ellis,	
   R.	
   J.	
   (2001)	
   Macromolecular	
   crowding:	
   an	
   important	
   but	
   neglected	
   aspect	
   of	
   the	
  
intracellular	
   environment	
   (vol	
   11,	
   pg	
   114,	
   2001),	
   Current	
   Opinion	
   in	
   Structural	
   Biology	
  
11,	
  500-­‐500.	
  
Minton,	
   A.	
   P.	
   (2000)	
   Implications	
   of	
   macromolecular	
   crowding	
   for	
   protein	
   assembly,	
  
Current	
  Opinion	
  in	
  Structural	
  Biology	
  10,	
  34-­‐39.	
  
O'Neill,	
   J.	
   W.,	
   Kim,	
   D.	
   E.,	
   Johnsen,	
   K.,	
   Baker,	
   D.,	
   and	
   Zhang,	
   K.	
   Y.	
   J.	
   (2001)	
   Single-­‐site	
  
mutations	
   induce	
   3D	
   domain	
   swapping	
   in	
   the	
   B1	
   domain	
   of	
   protein	
   L	
   from	
  
Peptostreptococcus	
  magnus,	
  Structure	
  9,	
  1017-­‐1027.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nature	
  chemical	
  biology	
  10,	
  884-­‐891.	
  
Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2004)	
  Monitoring	
  protein	
  stability	
  and	
  aggregation	
  in	
  
vivo	
  by	
  real-­‐time	
  fluorescent	
  labeling,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  101,	
  523-­‐528.	
  
182	
  

69.	
  
	
  
70.	
  
	
  
71.	
  
	
  
72.	
  
	
  
73.	
  
	
  
74.	
  
	
  
75.	
  
	
  
76.	
  
	
  
77.	
  
	
  
78.	
  
	
  
79.	
  
	
  
80.	
  

Ignatova,	
  Z.,	
  and	
  Gierasch,	
  L.	
  M.	
  (2006)	
  Inhibition	
  of	
  protein	
  aggregation	
  in	
  vitro	
  and	
  in	
  
vivo	
  by	
  a	
  natural	
  osmoprotectant,	
  Proc	
  Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  103,	
  13357-­‐13361.	
  
Ropson,	
  I.	
  J.,	
  Boyer,	
  J.	
  A.,	
  Schaeffer,	
  B.	
  A.,	
  and	
  Dalessio,	
  P.	
  M.	
  (2009)	
  Comparison	
  of	
  the	
  
folding	
   mechanism	
   of	
   highly	
   homologous	
   proteins	
   in	
   the	
   lipid-­‐binding	
   protein	
   family,	
  
Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  75,	
  799-­‐806.	
  
Burns,	
   L.	
   L.,	
   and	
   Ropson,	
   I.	
   J.	
   (2001)	
   Folding	
   of	
   intracellular	
   retinol	
   and	
   retinoic	
   acid	
  
binding	
  proteins,	
  Proteins:	
  Structure,	
  Function,	
  and	
  Bioinformatics	
  43,	
  292-­‐302.	
  
Gershenson,	
  A.,	
  Gierasch,	
  L.	
  M.,	
  Pastore,	
  A.,	
  and	
  Radford,	
  S.	
  E.	
  (2014)	
  Energy	
  landscapes	
  
of	
  functional	
  proteins	
  are	
  inherently	
  risky,	
  Nat.	
  Chem.	
  Biol.	
  10,	
  884-­‐891.	
  
Crist,	
   R.	
   M.,	
   Vasileiou,	
   C.,	
   Rabago-­‐Smith,	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2006)	
  
Engineering	
  a	
  Rhodopsin	
  Protein	
  Mimic,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  128,	
  
4522-­‐4523.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  J	
  Am	
  Chem	
  Soc	
  135,	
  16111-­‐16119.	
  
Wang,	
  W.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2014)	
  The	
  photochemical	
  determinants	
  of	
  color	
  
vision:	
   revealing	
   how	
   opsins	
   tune	
   their	
   chromophore's	
   absorption	
   wavelength,	
  
BioEssays	
  :	
  news	
  and	
  reviews	
  in	
  molecular,	
  cellular	
  and	
  developmental	
  biology	
  36,	
  65-­‐74.	
  
Wang,	
   W.,	
   Nossoni,	
   Z.,	
   Berbasova,	
   T.,	
   Watson,	
   C.	
   T.,	
   Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Vasileiou,	
   C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   electronic	
   absorption	
   of	
   protein-­‐
embedded	
  all-­‐trans-­‐retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
Nosrati,	
   M.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
   Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   H.	
   (2016)	
   A	
  
Photoisomerizing	
  Rhodopsin	
  Mimic	
  Observed	
  at	
  Atomic	
  Resolution,	
  J	
  Am	
  Chem	
  Soc	
  138,	
  
8802-­‐8808.	
  
Berbasova,	
   T.,	
   Santos,	
   E.	
   M.,	
   Nosrati,	
   M.,	
   Vasileiou,	
   C.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
  
(2016)	
  Light-­‐Activated	
  Reversible	
  Imine	
  Isomerization:	
  Towards	
  a	
  Photochromic	
  Protein	
  
Switch,	
  Chembiochem	
  :	
  a	
  European	
  journal	
  of	
  chemical	
  biology	
  17,	
  407-­‐414.	
  
Yapici,	
   I.,	
   Lee,	
   K.	
   S.,	
   Berbasova,	
   T.,	
   Nosrati,	
   M.,	
   Jia,	
   X.,	
   Vasileiou,	
   C.,	
   Wang,	
   W.,	
   Santos,	
   E.	
  
M.,	
  Geiger,	
  J.	
  H.,	
  and	
  Borhan,	
  B.	
  (2015)	
  "Turn-­‐on"	
  protein	
  fluorescence:	
  in	
  situ	
  formation	
  
of	
  cyanine	
  dyes,	
  J	
  Am	
  Chem	
  Soc	
  137,	
  1073-­‐1080.	
  
Wang,	
  W.,	
  Nossoni,	
  Z.,	
  Berbasova,	
  T.,	
  Watson,	
  C.	
  T.,	
  Yapici,	
  I.,	
  Lee,	
  K.	
  S.	
  S.,	
  Vasileiou,	
  C.,	
  
Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2012)	
   Tuning	
   the	
   Electronic	
   Absorption	
   of	
   Protein-­‐
Embedded	
  All-­‐trans-­‐Retinal,	
  Science	
  338,	
  1340-­‐1343.	
  
183	
  

81.	
  
	
  
82.	
  

	
  
83.	
  
	
  
84.	
  

	
  
85.	
  
	
  
86.	
  
	
  
87.	
  
	
  
88.	
  
	
  
89.	
  
	
  
90.	
  
	
  
91.	
  

Gunasekaran,	
   K.,	
   Hagler,	
   A.	
   T.,	
   and	
   Gierasch,	
   L.	
   M.	
   (2004)	
   Sequence	
   and	
   structural	
  
analysis	
   of	
   cellular	
   retinoic	
   acid-­‐binding	
   proteins	
   reveals	
   a	
   network	
   of	
   conserved	
  
hydrophobic	
  interactions,	
  Proteins-­‐Structure	
  Function	
  and	
  Genetics	
  54,	
  179-­‐194.	
  
Nossoni,	
   Z.,	
   Assar,	
   Z.,	
   Yapici,	
   I.,	
   Nosrati,	
   M.,	
   Wang,	
   W.,	
   Berbasova,	
   T.,	
   Vasileiou,	
   C.,	
  
Borhan,	
   B.,	
   and	
   Geiger,	
   J.	
   (2014)	
   Structures	
   of	
   holo	
   wild-­‐type	
   human	
   cellular	
   retinol-­‐
binding	
  protein	
  II	
  (hCRBPII)	
  bound	
  to	
  retinol	
  and	
  retinal,	
  Acta	
  Crystallogr.	
  Sect.	
  D.	
  Biol.	
  
Crystallogr.	
  70,	
  3226-­‐3232.	
  
Shameer,	
   K.,	
   Shingate,	
   P.	
   N.,	
   Manjunath,	
   S.	
   C.	
   P.,	
   Karthika,	
   M.,	
   Pugalenthi,	
   G.,	
   and	
  
Sowdhamini,	
   R.	
   (2011)	
   3DSwap:	
   curated	
   knowledgebase	
   of	
   proteins	
   involved	
   in	
   3D	
  
domain	
  swapping,	
  Database-­‐the	
  Journal	
  of	
  Biological	
  Databases	
  and	
  Curation.	
  
Sillitoe,	
   I.,	
   Cuff,	
   A.	
   L.,	
   Dessailly,	
   B.	
   H.,	
   Dawson,	
   N.	
   L.,	
   Furnham,	
   N.,	
   Lee,	
   D.,	
   Lees,	
   J.	
   G.,	
  
Lewis,	
   T.	
   E.,	
   Studer,	
   R.	
   A.,	
   Rentzsch,	
   R.,	
   Yeats,	
   C.,	
   Thornton,	
   J.	
   M.,	
   and	
   Orengo,	
   C.	
   A.	
  
(2013)	
  New	
  functional	
  families	
  (FunFams)	
  in	
  CATH	
  to	
  improve	
  the	
  mapping	
  of	
  conserved	
  
functional	
  sites	
  to	
  3D	
  structures,	
  Nucleic	
  Acids	
  Research	
  41,	
  D490-­‐D498.	
  
Englander,	
   S.	
   W.,	
   and	
   Mayne,	
   L.	
   (2014)	
   The	
   nature	
   of	
   protein	
   folding	
   pathways,	
   Proc	
  
Natl	
  Acad	
  Sci	
  U	
  S	
  A	
  111,	
  15873-­‐15880.	
  
Englander,	
  S.	
  W.,	
  Mayne,	
  L.,	
  and	
  Krishna,	
  M.	
  M.	
  (2007)	
  Protein	
  folding	
  and	
  misfolding:	
  
mechanism	
  and	
  principles,	
  Quarterly	
  reviews	
  of	
  biophysics	
  40,	
  287-­‐326.	
  
Krishna,	
   M.	
   M.,	
   and	
   Englander,	
   S.	
   W.	
   (2007)	
   A	
   unified	
   mechanism	
   for	
   protein	
   folding:	
  
predetermined	
   pathways	
   with	
   optional	
   errors,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  16,	
  449-­‐464.	
  
Mayne,	
   L.,	
   and	
   Englander,	
   S.	
   W.	
   (2000)	
   Two-­‐state	
   vs.	
   multistate	
   protein	
   unfolding	
  
studied	
   by	
   optical	
   melting	
   and	
   hydrogen	
   exchange,	
   Protein	
   science	
   :	
   a	
   publication	
   of	
   the	
  
Protein	
  Society	
  9,	
  1873-­‐1877.	
  
Roder,	
  H.,	
  Elove,	
  G.	
  A.,	
  and	
  Englander,	
  S.	
  W.	
  (1988)	
  Structural	
  characterization	
  of	
  folding	
  
intermediates	
   in	
   cytochrome	
   c	
   by	
   H-­‐exchange	
   labelling	
   and	
   proton	
   NMR,	
   Nature	
   335,	
  
700-­‐704.	
  
Martins,	
   N.	
   F.,	
   and	
   Santoro,	
   M.	
   M.	
   (1999)	
   Partially	
   folded	
   intermediates	
   during	
  
trypsinogen	
  denaturation,	
  Brazilian	
  journal	
  of	
  medical	
  and	
  biological	
  research	
  =	
  Revista	
  
brasileira	
  de	
  pesquisas	
  medicas	
  e	
  biologicas	
  32,	
  673-­‐682.	
  
Medina,	
   E.,	
   Cordova,	
   C.,	
   Villalobos,	
   P.,	
   Reyes,	
   J.,	
   Komives,	
   E.	
   A.,	
   Ramirez-­‐Sarmiento,	
   C.	
  
A.,	
   and	
   Babul,	
   J.	
   (2016)	
   Three-­‐Dimensional	
   Domain	
   Swapping	
   Changes	
   the	
   Folding	
  
Mechanism	
  of	
  the	
  Forkhead	
  Domain	
  of	
  FoxP1,	
  Biophysical	
  journal	
  110,	
  2349-­‐2360.	
  

	
  
184	
  

92.	
  
	
  
93.	
  
	
  
94.	
  
	
  
95.	
  

	
  
96.	
  
	
  
97.	
  
	
  
98.	
  

	
  
99.	
  
	
  
100.	
  

	
  
101.	
  
	
  
102.	
  

Zwanzig,	
   R.	
   (1997)	
   Two-­‐state	
   models	
   of	
   protein	
   folding	
   kinetics,	
   Proceedings	
   of	
   the	
  
National	
  Academy	
  of	
  Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  94,	
  148-­‐150.	
  
Sanson,	
   B.,	
   Wang,	
   T.,	
   Sun,	
   J.,	
   Wang,	
   L.,	
   Kaczocha,	
   M.,	
   Ojima,	
   I.,	
   Deutsch,	
   D.,	
   and	
   Li,	
   H.	
  
(2014)	
  Crystallographic	
  study	
  of	
  FABP5	
  as	
  an	
  intracellular	
  endocannabinoid	
  transporter,	
  
Acta	
  crystallographica.	
  Section	
  D,	
  Biological	
  crystallography	
  70,	
  290-­‐298.	
  
Assar,	
   Z.,	
   Nossoni,	
   Z.,	
   Wang,	
   W.,	
   Santos,	
   Elizabeth	
  M.,	
   Kramer,	
   K.,	
   McCornack,	
   C.,	
  
Vasileiou,	
  C.,	
  Borhan,	
  B.,	
  and	
  Geiger,	
  James	
  H.	
  Domain-­‐Swapped	
  Dimers	
  of	
  Intracellular	
  
Lipid-­‐Binding	
  Proteins:	
  Evidence	
  for	
  Ordered	
  Folding	
  Intermediates,	
  Structure.	
  
Silvaroli,	
   J.	
   A.,	
   Arne,	
   J.	
   M.,	
   Chelstowska,	
   S.,	
   Kiser,	
   P.	
   D.,	
   Banerjee,	
   S.,	
   and	
   Golczak,	
   M.	
  
(2016)	
   Ligand	
   Binding	
   Induces	
   Conformational	
   Changes	
   in	
   Human	
   Cellular	
   Retinol-­‐
binding	
  Protein	
  1	
  (CRBP1)	
  Revealed	
  by	
  Atomic	
  Resolution	
  Crystal	
  Structures,	
  The	
  Journal	
  
of	
  biological	
  chemistry	
  291,	
  8528-­‐8540.	
  
Gillilan,	
  R.	
  E.,	
  Ayers,	
  S.	
  D.,	
  and	
  Noy,	
  N.	
  (2007)	
  Structural	
  basis	
  for	
  activation	
  of	
  fatty	
  acid-­‐
binding	
  protein	
  4,	
  J	
  Mol	
  Biol	
  372,	
  1246-­‐1260.	
  
Sessler,	
  R.	
  J.,	
  and	
  Noy,	
  N.	
  (2005)	
  A	
  ligand-­‐activated	
  nuclear	
  localization	
  signal	
  in	
  cellular	
  
retinoic	
  acid	
  binding	
  protein-­‐II,	
  Molecular	
  cell	
  18,	
  343-­‐353.	
  
Armstrong,	
  E.	
  H.,	
  Goswami,	
  D.,	
  Griffin,	
  P.	
  R.,	
  Noy,	
  N.,	
  and	
  Ortlund,	
  E.	
  A.	
  (2014)	
  Structural	
  
Basis	
   for	
   Ligand	
   Regulation	
   of	
   the	
   Fatty	
   Acid-­‐binding	
   Protein	
   5,	
   Peroxisome	
   Proliferator-­‐
activated	
   Receptor	
   β/δ	
   (FABP5-­‐PPARβ/δ)	
   Signaling	
   Pathway,	
   The	
   Journal	
   of	
   biological	
  
chemistry	
  289,	
  14941-­‐14954.	
  
Greenfield,	
  N.	
  J.	
  (2006)	
  Using	
  circular	
  dichroism	
  collected	
  as	
  a	
  function	
  of	
  temperature	
  
to	
  determine	
  the	
  thermodynamics	
  of	
  protein	
  unfolding	
  and	
  binding	
  interactions,	
  Nature	
  
Protocols	
  1,	
  2527-­‐2535.	
  
Ramos,	
   C.	
   R.	
   R.,	
   Oyama,	
   S.,	
   Jr.,	
   Sforca,	
   M.	
   L.,	
   Pertinhez,	
   T.	
   A.,	
   Ho,	
   P.	
   L.,	
   and	
   Spisni,	
   A.	
  
(2008)	
   2POA:	
   Schistosoma	
   mansoni	
   Sm14	
   Fatty	
   Acid-­‐Binding	
   Protein:	
   improvement	
   of	
  
protein	
   stability	
   by	
   substitution	
   of	
   the	
   single	
   Cys62	
   residue,	
   Worldwide	
   Protein	
   Data	
  
Bank.	
  
Gill,	
   S.	
   C.,	
   and	
   Vonhippel,	
   P.	
   H.	
   (1989)	
   CALCULATION	
   OF	
   PROTEIN	
   EXTINCTION	
  
COEFFICIENTS	
   FROM	
   AMINO-­‐ACID	
   SEQUENCE	
   DATA,	
   Analytical	
   Biochemistry	
   182,	
   319-­‐
326.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Davis,	
  I.	
  W.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  
J.,	
   Hung,	
   L.	
   W.,	
   Kapral,	
   G.	
   J.,	
   Grosse-­‐Kunstleve,	
   R.	
   W.,	
   McCoy,	
   A.	
   J.,	
   Moriarty,	
   N.	
   W.,	
  
Oeffner,	
  R.,	
  Read,	
  R.	
  J.,	
  Richardson,	
  D.	
  C.,	
  Richardson,	
  J.	
  S.,	
  Terwilliger,	
  T.	
  C.,	
  and	
  Zwart,	
  
P.	
   H.	
   (2010)	
   PHENIX:	
   a	
   comprehensive	
   Python-­‐based	
   system	
   for	
   macromolecular	
  
185	
  

	
  
103.	
  
	
  
104.	
  

	
  
105.	
  
	
  
106.	
  
	
  
107.	
  
	
  
108.	
  
	
  
109.	
  
	
  
110.	
  

	
  
111.	
  

	
  
112.	
  

structure	
  solution,	
  Acta	
  Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  66,	
  213-­‐
221.	
  
Bailey,	
  S.	
  (1994)	
  THE	
  CCP4	
  SUITE	
  -­‐	
  PROGRAMS	
  FOR	
  PROTEIN	
  CRYSTALLOGRAPHY,	
  Acta	
  
Crystallographica	
  Section	
  D-­‐Biological	
  Crystallography	
  50,	
  760-­‐763.	
  
Adams,	
  P.	
  D.,	
  Afonine,	
  P.	
  V.,	
  Bunkoczi,	
  G.,	
  Chen,	
  V.	
  B.,	
  Echols,	
  N.,	
  Headd,	
  J.	
  J.,	
  Hung,	
  L.-­‐
W.,	
  Jain,	
  S.,	
  Kapral,	
  G.	
  J.,	
  Kunstleve,	
  R.	
  W.	
  G.,	
  McCoy,	
  A.	
  J.,	
  Moriarty,	
  N.	
  W.,	
  Oeffner,	
  R.	
  
D.,	
   Read,	
   R.	
   J.,	
   Richardson,	
   D.	
   C.,	
   Richardson,	
   J.	
   S.,	
   Terwilliger,	
   T.	
   C.,	
   and	
   Zwart,	
   P.	
   H.	
  
(2011)	
  The	
  Phenix	
  software	
  for	
  automated	
  determination	
  of	
  macromolecular	
  structures,	
  
Methods	
  55,	
  94-­‐106.	
  
Murshudov,	
  G.	
  N.,	
  Vagin,	
  A.	
  A.,	
  and	
  Dodson,	
  E.	
  J.	
  (1997)	
  Refinement	
  of	
  macromolecular	
  
structures	
   by	
   the	
   maximum-­‐likelihood	
   method,	
   Acta	
   Crystallographica	
   Section	
   D-­‐
Biological	
  Crystallography	
  53,	
  240-­‐255.	
  
Ong,	
   D.	
   E.,	
   and	
   Page,	
   D.	
   L.	
   (1987)	
   Cellular	
   retinol-­‐binding	
   protein	
   (type	
   two)	
   is	
   abundant	
  
in	
  human	
  small	
  intestine,	
  Journal	
  of	
  Lipid	
  Research	
  28,	
  739-­‐745.	
  
Haunerland,	
   N.	
   H.,	
   and	
   Spener,	
   F.	
   (2004)	
   Fatty	
   acid-­‐binding	
   proteins	
   –	
   insights	
   from	
  
genetic	
  manipulations,	
  Progress	
  in	
  Lipid	
  Research	
  43,	
  328-­‐349.	
  
Gillilan,	
   R.	
   E.,	
   Ayers,	
   S.	
   D.,	
   and	
   Noy,	
   N.	
   (2007)	
   Structural	
   Basis	
   for	
   Activation	
   of	
   Fatty	
  
Acid-­‐binding	
  Protein	
  4,	
  Journal	
  of	
  Molecular	
  Biology	
  372,	
  1246-­‐1260.	
  
Kaczocha,	
   M.,	
   Glaser,	
   S.	
   T.,	
   and	
   Deutsch,	
   D.	
   G.	
   (2009)	
   Identification	
   of	
   intracellular	
  
carriers	
  for	
  the	
  endocannabinoid	
  anandamide,	
  Proceedings	
  of	
  the	
  National	
  Academy	
  of	
  
Sciences	
  of	
  the	
  United	
  States	
  of	
  America	
  106,	
  6375-­‐6380.	
  
Sulsky,	
  R.,	
  Magnin,	
  D.	
  R.,	
  Huanga,	
  Y.,	
  Simpkins,	
  L.,	
  Taunk,	
  P.,	
  Patel,	
  M.,	
  Zhu,	
  Y.,	
  Stouch,	
  T.	
  
R.,	
  Bassolino-­‐Klimas,	
  D.,	
  Parker,	
  R.,	
  Harrity,	
  T.,	
  Stoffel,	
  R.,	
  Taylor,	
  D.	
  S.,	
  Lavoie,	
  T.	
  B.,	
  Kish,	
  
K.,	
  Jacobson,	
  B.	
  L.,	
  Sheriff,	
  S.,	
  Adam,	
  L.	
  P.,	
  Ewing,	
  W.	
  R.,	
  and	
  Robl,	
  J.	
  A.	
  (2007)	
  Potent	
  and	
  
selective	
   biphenyl	
   azole	
   inhibitors	
   of	
   adipocyte	
   fatty	
   acid	
   binding	
   pirotein	
   (aFABP),	
  
Bioorganic	
  &	
  Medicinal	
  Chemistry	
  Letters	
  17,	
  3511-­‐3515.	
  
Yapici,	
   I.,	
   Lee,	
   K.	
   S.	
   S.,	
   Berbasova,	
   T.,	
   Nosrati,	
   M.,	
   Jia,	
   X.	
   F.,	
   Vasileiou,	
   C.,	
   Wang,	
   W.	
   J.,	
  
Santos,	
   E.	
   M.,	
   Geiger,	
   J.	
   H.,	
   and	
   Borhan,	
   B.	
   (2015)	
   "Turn-­‐On"	
   Protein	
   Fluorescence:	
   In	
  
Situ	
   Formation	
   of	
   Cyanine	
   Dyes,	
   Journal	
   of	
   the	
   American	
   Chemical	
   Society	
   137,	
   1073-­‐
1080.	
  
Berbasova,	
  T.,	
  Nosrati,	
  M.,	
  Vasileiou,	
  C.,	
  Wang,	
  W.,	
  Lee,	
  K.	
  S.,	
  Yapici,	
  I.,	
  Geiger,	
  J.	
  H.,	
  and	
  
Borhan,	
  B.	
  (2013)	
  Rational	
  design	
  of	
  a	
  colorimetric	
  pH	
  sensor	
  from	
  a	
  soluble	
  retinoic	
  acid	
  
chaperone,	
  Journal	
  of	
  the	
  American	
  Chemical	
  Society	
  135,	
  16111-­‐16119.	
  

	
  
186	
  

113.	
  
	
  
114.	
  
	
  
115.	
  
	
  
116.	
  
	
  
117.	
  

	
  
118.	
  

	
  
119.	
  
	
  
120.	
  
	
  
121.	
  

	
  
122.	
  
	
  

Abraham,	
  M.	
  H.,	
  Grellier,	
  P.	
  L.,	
  Abboud,	
  J.	
  L.	
  M.,	
  Doherty,	
  R.	
  M.,	
  and	
  Taft,	
  R.	
  W.	
  (1988)	
  
Solvent	
   Effects	
   in	
   Organic-­‐Chemistry	
   -­‐	
   Recent	
   Developments,	
   Canadian	
   Journal	
   of	
  
Chemistry-­‐Revue	
  Canadienne	
  De	
  Chimie	
  66,	
  2673-­‐2686.	
  
Ando,	
   Y.,	
   Homma,	
   Y.,	
   Hiruta,	
   Y.,	
   Citterio,	
   D.,	
   and	
   Suzuki,	
   K.	
   (2009)	
   Structural	
  
characteristics	
   and	
   optical	
   properties	
   of	
   a	
   series	
   of	
   solvatochromic	
   fluorescent	
   dyes	
  
displaying	
  long-­‐wavelength	
  emission,	
  Dyes	
  and	
  Pigments	
  83,	
  198-­‐206.	
  
Carlotti,	
   B.,	
   Flamini,	
   R.,	
   Kikas,	
   I.,	
   Mazzucato,	
   U.,	
   and	
   Spalletti,	
   A.	
   (2012)	
   Intramolecular	
  
charge	
   transfer,	
   solvatochromism	
   and	
   hyperpolarizability	
   of	
   compounds	
   bearing	
  
ethenylene	
  or	
  ethynylene	
  bridges,	
  Chemical	
  Physics	
  407,	
  9-­‐19.	
  
Kamlet,	
  M.	
  J.,	
  and	
  Taft,	
  R.	
  W.	
  (1976)	
  Solvatochromic	
  Comparison	
  Method	
  1.	
  Beta-­‐Scale	
  
of	
  Solvent	
  Hydrogen-­‐Bond	
  Acceptor	
  (Hba)	
  Basicities,	
  Journal	
  of	
  the	
  American	
  Chemical	
  
Society	
  98,	
  377-­‐383.	
  
Yutani,	
   K.,	
   Ogasahara,	
   K.,	
   Tsujita,	
   T.,	
   and	
   Sugino,	
   Y.	
   (1987)	
   Dependence	
   of	
  
Conformational	
   Stability	
   on	
   Hydrophobicity	
   of	
   the	
   Amino-­‐Acid	
   Residue	
   in	
   a	
   Series	
   of	
  
Variant	
   Proteins	
   Substituted	
   at	
   a	
   Unique	
   Position	
   of	
   Tryptophan	
   Synthase	
   Alpha-­‐
Subunit,	
   Proceedings	
   of	
   the	
   National	
   Academy	
   of	
   Sciences	
   of	
   the	
   United	
   States	
   of	
  
America	
  84,	
  4441-­‐4444.	
  
Noguchi,	
  J.,	
  Chaen,	
  K.,	
  Nhuan	
  Thi,	
  V.,	
  Akasaka,	
  T.,	
  Shimada,	
  H.,	
  Nakashima,	
  T.,	
  Nishi,	
  A.,	
  
Satoh,	
  H.,	
  Omori,	
  T.,	
  Kakuta,	
  Y.,	
  and	
  Kimura,	
  M.	
  (2011)	
  Crystal	
  structure	
  of	
  the	
  branching	
  
enzyme	
  I	
  (BEI)	
  from	
  Oryza	
  sativa	
  L	
  with	
  implications	
  for	
  catalysis	
  and	
  substrate	
  binding,	
  
Glycobiology	
  21,	
  1108-­‐1116.	
  
Feng,	
   L.,	
   Fawaz,	
   R.,	
   Hovde,	
   S.,	
   Gilbert,	
   L.,	
   Chiou,	
   J.,	
   and	
   Geiger,	
   J.	
   H.	
   (2015)	
   Crystal	
  
Structures	
   of	
   Escherichia	
   coli	
   Branching	
   Enzyme	
   in	
   Complex	
   with	
   Linear	
  
Oligosaccharides,	
  Biochemistry	
  54,	
  6207-­‐6218.	
  
Henrissat,	
  B.	
  (1991)	
  A	
  classification	
  of	
  glycosyl	
  hydrolases	
  based	
  on	
  amino	
  acid	
  sequence	
  
similarities.,	
  Biochemical	
  Journal	
  280,	
  309-­‐316.	
  
Jespersen,	
  H.	
  M.,	
  MacGregor,	
  E.	
  A.,	
  Henrissat,	
  B.,	
  Sierks,	
  M.	
  R.,	
  and	
  Sevensson,	
  B.	
  (1993)	
  
Starch-­‐	
   and	
   Glycogen-­‐Debranching	
   and	
   Branching	
   Enzymes:	
   Prediction	
   of	
   Structural	
  
Features	
   of	
   the	
   Catalytic	
   (α/β)8-­‐Barrel	
   Domain	
   and	
   Evolutionary	
   Relationship	
   to	
   Other	
  
Amylolytic	
  Enzymes,	
  Journal	
  of	
  Protein	
  Chemistry	
  12,	
  791-­‐805.	
  
Baba,	
   T.,	
   Kimura,	
   K.,	
   Mizuno,	
   K.,	
   Etoh,	
   H.,	
   Ishida,	
   Y.,	
   Shida,	
   O.,	
   and	
   Arai,	
   Y.	
   (1991)	
  
SEQUENCE	
   CONSERVATION	
   OF	
   THE	
   CATALYTIC	
   REGIONS	
   OF	
   AMYLOLYTIC	
   ENZYMES	
   IN	
  
MAIZE	
  BRANCHING	
  ENZYME-­‐I,	
  Biochem.	
  Biophys.	
  Res.	
  Commun.	
  181,	
  87-­‐94.	
  

187	
  

123.	
  
	
  
124.	
  
	
  
125.	
  
	
  
126.	
  
	
  
127.	
  
	
  
128.	
  
	
  
129.	
  
	
  
130.	
  
	
  
131.	
  
	
  
132.	
  
	
  
133.	
  

	
  
134.	
  

Boyer,	
   C.	
   D.,	
   and	
   Preiss,	
   J.	
   (1981)	
   Evidence	
   for	
   Independent	
   Genetic	
   Control	
   of	
   the	
  
Multiple	
   Forms	
   of	
   Maize	
   Endosperm	
   Branching	
   Enzymes	
   and	
   Starch	
   Synthases,	
   Plant	
  
Physicology	
  67,	
  1141-­‐1145.	
  
Gao,	
   M.,	
   Fisher,	
   D.	
   K.,	
   Kim,	
   K.	
   N.,	
   Shannon,	
   J.	
   C.,	
   and	
   Guiltinan,	
   M.	
   J.	
   (1996)	
   Evolutionary	
  
conservation	
  and	
  expression	
  patterns	
  of	
  maize	
  starch	
  branching	
  enzyme	
  I	
  and	
  llb	
  genes	
  
suggests	
  isoform	
  specialization,	
  Plant	
  Molecular	
  Biology	
  30,	
  1223-­‐1232.	
  
Guan,	
   H.,	
   Li,	
   P.,	
   Imparl-­‐Radosevich,	
   J.,	
   Preiss,	
   J.,	
   and	
   Keeling,	
   P.	
   (1997)	
   Comparing	
   the	
  
Properties	
  of	
  Escherichia	
  coli	
  Branching	
  Enzyme	
  and	
  Maize	
  Branching	
  Enzyme,	
  Archives	
  
of	
  biochemistry	
  and	
  biophysics	
  342,	
  92-­‐98.	
  
Guan,	
   H.	
   P.,	
   Baba,	
   T.,	
   and	
   Preiss,	
   J.	
   (1994)	
   Expression	
   of	
   Branching	
   Enzyme	
   I	
   of	
   Maize	
  
Endosperm	
  in	
  Escherichia	
  coli,	
  Plant	
  Physicology	
  104,	
  1449-­‐1453.	
  
Guan,	
   H.	
   P.,	
   Kuriki,	
   T.,	
   Sivak,	
   M.,	
   and	
   Preiss,	
   J.	
   (1995)	
   Maize	
   Branching	
   Enzyme	
   Catalyzes	
  
synthesis	
   of	
   Glycogen-­‐like	
   Polysaccharide	
   in	
   GlgB-­‐Deficient	
   Escherichia	
   Coli,	
   Proceedings	
  
of	
  the	
  National	
  Academy	
  of	
  Sciences	
  in	
  the	
  United	
  States	
  of	
  America	
  92,	
  964-­‐967.	
  
Guan,	
   H.	
   P.,	
   and	
   Preiss,	
   J.	
   (1993)	
   Differentiation	
   of	
   the	
   Properties	
   of	
   the	
   Branching	
  
lsozymes	
  from	
  Maize	
  (Zea	
  mays),	
  Plant	
  Physiology	
  102,	
  1269-­‐1273.	
  
Kuriki,	
   T.,	
   Guan,	
   H.,	
   Sivak,	
   M.,	
   and	
   Preiss,	
   J.	
   (1996)	
   Analysis	
   of	
   the	
   Active	
   Center	
   of	
  
Branching	
  Enzyme	
  II	
  from	
  Maize	
  Endosperm,	
  J	
  Protein	
  Chem	
  15,	
  305-­‐313.	
  
Kuriki,	
  T.,	
  Stewart,	
  D.	
  C.,	
  and	
  Preiss,	
  J.	
  (1997)	
  Construction	
  of	
  Chimeric	
  Enzymes	
  out	
  of	
  
Maize	
   Endosperm	
   Branching	
   Enzymes	
   I	
   and	
   II::	
   ACTIVITY	
   AND	
   PROPERTIES,	
   Journal	
   of	
  
Biological	
  Chemistry	
  272,	
  28999-­‐29004.	
  
Takeda,	
   Y.,	
   Guan,	
   H.-­‐P.,	
   and	
   Preiss,	
   J.	
   (1993)	
   Branching	
   of	
   amylose	
   by	
   the	
   branching	
  
isoenzymes	
  of	
  maize	
  endosperm,	
  Carbohhydr.	
  Res.	
  240,	
  253-­‐263.	
  
Kawasaki,	
   T.,	
   Mizumo,	
   K.,	
   Baba,	
   T.,	
   and	
   Shimada,	
   H.	
   (1993)	
   Molecular	
   analysis	
   of	
   the	
  
gene	
  encoding	
  a	
  rice	
  starch	
  branching	
  enzyme,	
  Molecular	
  and	
  General	
  Genetics	
  237,	
  10-­‐
16.	
  
Mizuno,	
  K.,	
  Kawasaki,	
  T.,	
  Shimada,	
  H.,	
  Satoh,	
  H.,	
  Kobayashi,	
  E.,	
  Okumura,	
  S.,	
  Arai,	
  Y.,	
  and	
  
Baba,	
  T.	
  (1993)	
  Alteration	
  of	
  the	
  Structural	
  Properties	
  of	
  Starch	
  Components	
  by	
  the	
  Lack	
  
of	
   an	
   Isoform	
   of	
   Starch	
   Branching	
   Enzyme	
   in	
   Rice	
   Seeds,	
   The	
   Journal	
   of	
   biological	
  
chemistry	
  268,	
  19084-­‐19091.	
  
Mizuno,	
   K.,	
   Kimura,	
   K.,	
   Arai,	
   Y.,	
   Kawasaki,	
   T.,	
   Shimada,	
   H.,	
   and	
   Baba,	
   T.	
   (1992)	
   Starch	
  
Branching	
  Enzymes	
  from	
  Immature	
  Rice	
  Seeds,	
  Journal	
  of	
  Biochemistry	
  112,	
  643-­‐651.	
  

	
  
188	
  

135.	
  
	
  
136.	
  
	
  
137.	
  
	
  
138.	
  
	
  
139.	
  
	
  
140.	
  
	
  
141.	
  

	
  
142.	
  
	
  
143.	
  

	
  
144.	
  
	
  

Mizuno,	
   K.,	
   Kobayashi,	
   E.,	
   Tachibana,	
   M.,	
   Tsutomu,	
   K.,	
   Fujimura,	
   T.,	
   Funane,	
   K.,	
  
Kobayashi,	
   M.,	
   and	
   Baba,	
   T.	
   (2001)	
   Characterization	
   of	
   an	
   isoform	
   of	
   rice	
   starch	
  
branching	
  enzyme,	
  RBE4,	
  in	
  developing	
  seeds,	
  Plant	
  Cell	
  Physiol.	
  42,	
  349-­‐357.	
  
Nakamura,	
  Y.,	
  Takeichi,	
  T.,	
  Kawaguchi,	
  K.,	
  and	
  Yamanouchi,	
  H.	
  (1992)	
  Purification	
  of	
  two	
  
forms	
   of	
   starch	
   branching	
   enzyme	
   (Q-­‐enzyme)	
   from	
   developing	
   rice	
   endosperm,	
   Plant	
  
Physiology	
  84,	
  329	
  –335.	
  
Nakamura,	
  Y.,	
  Utsumi,	
  Y.,	
  Sawada,	
  T.,	
  Aihara,	
  S.,	
  Utsumi,	
  C.,	
  Yoshida,	
  M.,	
  and	
  Kitamura,	
  
S.	
   (2010)	
   Characterization	
   of	
   the	
   reactions	
   of	
   starch	
   branching	
   enzymes	
   from	
   rice	
  
endosperm,	
  Plant	
  and	
  Cell	
  Physiology	
  51,	
  776-­‐794.	
  
Morell,	
   M.	
   K.,	
   Blennow,	
   A.,	
   Kosar-­‐Hashemi,	
   B.,	
   and	
   Samuel,	
   M.	
   (1997)	
   Differential	
  
Expression	
   and	
   Properties	
   of	
   Starch	
   Branching	
   Enzyme	
   lsoforms	
   in	
   Developing	
   Wheat	
  
Endosperm,	
  Plant	
  Physiol.	
  113,	
  201-­‐208.	
  
Rahman,	
   S.,	
   Li,	
   Z.,	
   Abrahams,	
   S.,	
   Abbott,	
   D.,	
   Appels,	
   R.,	
   and	
   Morell,	
   M.	
   K.	
   (1999)	
  
Characterisation	
   of	
   a	
   gene	
   encoding	
   wheat	
   endosperm	
   starch	
   branching	
   enzyme-­‐I,	
  
Theor.	
  Appl.	
  Genet.	
  98,	
  156-­‐163.	
  
Repellin,	
  A.,	
  Büga,	
  M.,	
  and	
  Chibbar,	
  R.	
  N.	
  (2001)	
  Characterization	
  of	
  a	
  cDNA	
  encoding	
  a	
  
type	
   I	
   starch	
   branching	
   enzyme	
   produced	
   in	
   developing	
   wheat	
   (Triticum	
   aestivum	
  
L.)kernels,	
  Journal	
  of	
  Plant	
  Physiology	
  158,	
  91-­‐100.	
  
Sadequr	
   Rahman*,	
   A.	
   R.,	
   Zhongyi	
   Li,	
   Yasuhiko	
   Mukai,	
   Maki	
   Yamamoto,	
   Behjat	
   Kosar-­‐
Hashemi,	
   Sharon	
   Abrahams,	
   and	
   Matthew	
   K.	
   Morell.	
   (2001)	
   Comparison	
   of	
   Starch-­‐
Branching	
   Enzyme	
   Genes	
   Reveals	
   Evolutionary	
   Relationships	
   Among	
   Isoforms.	
  
Characterization	
  of	
  a	
  Gene	
  for	
  Starch-­‐Branching	
  Enzyme	
  IIa	
  from	
  the	
  Wheat	
  D	
  Genome	
  
Donor	
  Aegilops	
  tauschii1,	
  Plant	
  Physiology	
  125.	
  
Sun,	
  C.,	
  Sathish,	
  P.,	
  Ahlandsberg,	
  S.,	
  Deiber,	
  A.,	
  and	
  Jansson,	
  C.	
  (1997)	
  Identification	
  of	
  
four	
   starch-­‐branching	
   enzymes	
   in	
   barley	
   endosperm-­‐	
   partial	
   purification	
   of	
   forms	
   I,	
   Ila	
  
and	
  IIb,	
  New	
  Phytol.	
  137,	
  215-­‐222.	
  
ChuanXin,	
   S.,	
   Sathish,	
   P.,	
   Deiber,	
   A.,	
   and	
   Jansson,	
   C.	
   (1995)	
   Multiple	
   starch	
   branching	
  
enzymes	
   in	
   barley	
   endosperm,	
   Photosynthesis:	
   from	
   light	
   to	
   biosphere	
   Volume	
   V.	
  
Proceedings	
  of	
  the	
  Xth	
  International	
  Photosynthesis	
  Congress,	
  Montpellier,	
  France,	
  317-­‐
320.	
  
Sun,	
   C.,	
   Sathish,	
   P.,	
   Ahlandsberg,	
   S.,	
   and	
   Jansson,	
   C.	
   (1998)	
   The	
   Two	
   Genes	
   Encoding	
  
Starch-­‐Branching	
   Enzymes	
   IIa	
   and	
   IIb	
   Are	
   Differentially	
   Expressed	
   in	
   Barley,	
   Plant	
  
Physiol.	
  118,	
  37–49.	
  

189	
  

145.	
  
	
  
146.	
  
	
  
147.	
  
	
  
148.	
  
149.	
  
	
  
150.	
  

	
  
151.	
  
	
  
152.	
  
	
  
153.	
  
	
  
154.	
  
	
  
155.	
  
	
  
156.	
  

Rydberg,	
   U.,	
   Andersson,	
   L.,	
   Andersson,	
   R.,	
   A	
  ̊man2,	
   P.,	
   and	
   Larsson,	
   H.	
   (2001)	
  
Comparison	
  of	
  starch	
  branching	
  enzyme	
  I	
  and	
  II	
  from	
  potato,	
  Eur.	
  J.	
  Biochem.	
  268,	
  6140-­‐
6145.	
  
Visko-­‐Nielsen,	
   A.,	
   and	
   Blennow,	
   A.	
   (1998)	
   Isolation	
   of	
   starch	
   branching	
   enzyme	
   I	
   from	
  
potato	
  using	
  gamma-­‐cyclodextrin	
  affinity	
  chromatography,	
  J.	
  Chromatogr.	
  800,	
  382-­‐385.	
  
Salehuzzaman,	
   S.	
   N.	
   I.	
   M.,	
   Jacobsen,	
   E.,	
   and	
   Visser,	
   R.	
   G.	
   F.	
   (1992)	
   Cloning,	
   partial	
  
sequencing	
   and	
   expression	
   of	
   a	
   cDNA	
   coding	
   for	
   branching	
   enzyme	
   in	
   cassava,	
   Plant	
  
Molecular	
  Biology	
  20,	
  809-­‐819.	
  
Amith,	
  A.	
  M.	
  (1988)	
  Major	
  differences	
  in	
  isoforms	
  of	
  starch-­‐branching	
  enzyme	
  between	
  
developing	
  embryos	
  of	
  round-­‐	
  and	
  wrinkled-­‐seeded	
  peas	
  (Pisum	
  sativum	
  L.),	
  Planta	
  175,	
  
270-­‐279.	
  
Tomlinson,	
  K.	
  L.,	
  Lloydt,	
  J.	
  R.,	
  and	
  Smith,	
  A.	
  M.	
  (1997)	
  Importance	
  of	
  isoforms	
  of	
  starch-­‐
branching	
  enzyme	
  in	
  determining	
  the	
  structure	
  of	
  starch	
  in	
  pea	
  leaves,	
  The	
  Plant	
  Journal	
  
11,	
  31-­‐43.	
  
Zhu,	
  Z.	
  P.,	
  and	
  Chen,	
  S.	
  Q.	
  (1992,	
  20-­‐24	
  July)	
  Starch	
  accumulation	
  and	
  branching	
  enzyme	
  
gene	
   expression	
   during	
   seed	
   development	
   of	
   soybean	
   and	
   pea,	
   Proceedings	
   of	
   the	
  
Fourth	
   International	
   Workshop	
   on	
   Seeds:	
   basic	
   and	
   applied	
   aspects	
   of	
   seed	
   biology,	
  
Angers,	
  France	
  1,	
  127-­‐133.	
  
Nakamura,	
   Y.	
   (2002)	
   Towards	
   a	
   better	
   understanding	
   of	
   the	
   metabolic	
   system	
   for	
  
amylopectin	
  biosynthesis	
  in	
  plants,	
  Plant	
  Cell	
  Physiol.	
  43,	
  718-­‐725.	
  
Chaen,	
  K.,	
  Noguchi,	
  J.,	
  Omori,	
  T.,	
  Kakuta,	
  Y.,	
  and	
  Kimura,	
  M.	
  (2012)	
  Crystal	
  structure	
  of	
  
the	
   rice	
   branching	
   enzyme	
   I	
   (BEI)	
   in	
   complex	
   with	
   maltopentaose,	
   Biochemical	
   and	
  
Biophysical	
  Research	
  Communications	
  424,	
  508-­‐511.	
  
Vu,	
  N.	
  T.,	
  Shimada,	
  H.,	
  Kakuta,	
  Y.,	
  Nakashima,	
  T.,	
  Ida,	
  H.,	
  Omori,	
  T.,	
  Nishi,	
  A.,	
  Satoh,	
  H.,	
  
and	
  Kimura,	
  M.	
  (2008)	
  Biochemical	
  and	
  Crystallographic	
  Characterization	
  of	
  the	
  Starch	
  
Branching	
  Enzyme	
  I	
  (BEI)	
  from	
  Oryz	
  sativa	
  L,	
  Biosci.	
  Biotechnol.	
  Biochem.	
  72,	
  2858-­‐2866.	
  
Boyer,	
   C.	
   D.,	
   and	
   Preiss,	
   J.	
   (1978)	
   MULTIPLE	
   FORMS	
   OF	
   (1,4)-­‐a-­‐D-­‐GLUCAN,	
   (1,4)-­‐a-­‐D-­‐
GLUCAN-­‐6-­‐	
   GLYCOSYL	
   TRANSFERASE	
   FROM	
   DEVELOPING	
   Zea	
   mays	
   L.	
   KERNELS,	
  
Carbohydrate	
  Research	
  61,	
  321-­‐334.	
  	
  
Hizukuri,	
  S.,	
  Takeda,	
  Y.,	
  Yasuda,	
  M.,	
  and	
  Suzuki,	
  A.	
  (1981)	
  MULTI-­‐BRANCHED	
  NATURE	
  OF	
  
AMYLOSE	
   AND	
   THE	
   ACTION	
   OF	
   DE-­‐	
   BRANCHING	
   ENZYMES,	
   Carbohydrate	
   Research	
   94,	
  
205-­‐213.	
  
Binderup,	
   K.,	
   and	
   Preiss,	
   J.	
   (1998)	
   Glutamate-­‐459	
   Is	
   Important	
   for	
   Escherichia	
   coli	
  
Branching	
  Enzyme	
  Activity,	
  Biochemistry	
  37,	
  9033-­‐9037.	
  
190	
  

154.	
  
	
  
155.	
  
	
  
156.	
  
	
  
156.	
  
	
  

	
  

Boyer,	
   C.	
   D.,	
   and	
   Preiss,	
   J.	
   (1978)	
   MULTIPLE	
   FORMS	
   OF	
   (1,4)-­‐a-­‐D-­‐GLUCAN,	
   (1,4)-­‐a-­‐D-­‐
GLUCAN-­‐6-­‐	
   GLYCOSYL	
   TRANSFERASE	
   FROM	
   DEVELOPING	
   Zea	
   mays	
   L.	
   KERNELS,	
  
Carbohydrate	
  Research	
  61,	
  321-­‐334.	
  	
  
Hizukuri,	
  S.,	
  Takeda,	
  Y.,	
  Yasuda,	
  M.,	
  and	
  Suzuki,	
  A.	
  (1981)	
  MULTI-­‐BRANCHED	
  NATURE	
  OF	
  
AMYLOSE	
   AND	
   THE	
   ACTION	
   OF	
   DE-­‐	
   BRANCHING	
   ENZYMES,	
   Carbohydrate	
   Research	
   94,	
  
205-­‐213.	
  
Binderup,	
   K.,	
   and	
   Preiss,	
   J.	
   (1998)	
   Glutamate-­‐459	
   Is	
   Important	
   for	
   Escherichia	
   coli	
  
Branching	
  Enzyme	
  Activity,	
  Biochemistry	
  37,	
  9033-­‐9037.	
  
Binderup,	
   K.,	
   and	
   Preiss,	
   J.	
   (1998)	
   Glutamate-­‐459	
   Is	
   Important	
   for	
   Escherichia	
   coli	
  
Branching	
  Enzyme	
  Activity,	
  Biochemistry	
  37,	
  9033-­‐9037.	
  

	
  

191